Transcript: From $5 to "Service is a Product": How Shai Shperling Pushed Fiverr's Boldest Product Move

Eyal David: [00:00] So today we're going to talk about how you actually push, how you push a strategic initiative inside a company. Shaike is here, Shai Shperling from Empathy, and he came to talk about this, so let me briefly introduce Shai and myself — I'll skip myself, you can go look me up on LinkedIn — but Shaike, we actually worked together at Fiverr, today he's CPO at Empathy and he was there for 4 years, which is unusual these days, right? And basically at Fiverr you were the Head of Product, and you came here to talk specifically about an initiative from a long time ago that actually opened Fiverr up from the five dollars outward. Do you maybe want to say a word about yourself in half a sentence and then talk about why you're here?

Shai Shefrank: [01:15] Yeah, so, Shai, first of all thanks for the invitation, it's fun to be here again. I'm here because you invited me, and because we wanted to talk about interesting product things, and... Yeah, today at Empathy, before that at Fiverr, before that an entrepreneur, I advised a lot of companies, all kinds of companies, I founded, as I said, a few companies, some still exist and some no longer do. And honestly, when you reached out and said, let's talk about Fiverr, at first I thought, why? Like, why would anyone be interested in hearing what we did ten years ago about a product that's no longer new or interesting. And then suddenly — even though it is interesting, and not that it isn't — this thing is, it's old news as they say, it's a podcast, we're doing history, let's call it a product podcast. But yeah, suddenly when I thought about it in my head, what is there to say, there's a ton to say about this project. And I think, in terms of product, the place of product in a company, and changes and the evolution of a company, and strategy, and so many things that are in there, this is maybe one of the most interesting and complex projects I've touched, and I think there's a lot to talk about and it'll be really interesting.

Eyal David: [02:31] Great, so I think we'll dive into the agenda — oh, just before the agenda. Everywhere in the slides we'll show here, there's a QR code; anyone who isn't a member of our WhatsApp group, you're very welcome to join the community, and we'll add the link there to the recording and everything. Beyond that, as we go, we'd love it if you have any questions, you can also ask anonymously. At the end we'll bombard Shai with lots and lots of questions, so come on, let's start. Let's start — so basically our agenda will be split into something like 50 minutes, we'll talk, an open conversation really, and we'll leave time for questions of one kind or another if there are any. So let's actually start with a survey — you can pop it up — and we'll try to explain the essence here. We're doing a bit of expectation-setting; one sec, we want to know who you are, what role you're employed in, what you'd want to learn from this session — this is really at the level where we give you a minute, but it's really an American-style choice, and also how well you know Fiverr, so we know where to start from, okay? So we've started? Let's say. Let me take a moment. Call it... Chaim, there's also — it's worth saying that we're in the old Fiverr building, this is where everything actually happened.

Shai Shefrank: [03:40] True, true, a lot of... this project happened here, really really probably somewhere around here. Maybe it even paved the way to the new building, carried it forward. Could be, yeah, I understand so, among other things. Yeah, that was an investment later, I remember. Right.

Eyal David: [03:53] We, I think, did, after that — let's, a bit, let me go back a bit, go back to the history. First of all, it's exactly 10 years, this is November 2015.

Shai Shefrank: [04:01] 2015, and 9 years. And yeah, it was exactly together with the fundraising round. So that's it, this thing was without a doubt a serious driver for the business. Right? Yeah, amazing. It happened, we actually started — I started at Fiverr in Bnei Yamina... Well, you were at... with the two product people I work with today at Empathy, Likhai Gersht and Noga Nitzan. They were even before me at Fiverr in product, and I have the luck, the honor, to work with them today at Empathy too. Who else did you actually work with at Fiverr on this project? I worked with Asaf Porat, also a product person, a brilliant guy, I think — not much to say, he's an interesting guy. He was really my partner in crime on this project. I think this project, I think it was too stressful a project for anyone to take on by themselves and carry it, and the fact that Asaf was there to make me dare, and I was there to make him dare, made it happen. We'll talk about this project in the context of worries and fears, and I think that's... what an important theme, that product people... Whether their own worries or the worries of other people in the organization — but a lot of people worked on this. Wait, right, okay, are we done or do we give it another moment?

Eyal David: [05:17] There are five more on them. Oh, so let's give them the engagement. Tell us meanwhile, are you hiring now at Empathy, what does it look like?

Shai Shefrank: [05:25] We're hiring a bit in product specifically — right now, not at the moment, we just recruited for growth, but... growth? But yeah, yeah, growth, it's not exactly a product manager, maybe it's a track like that? It's not a track, it's a role, and a role. From the worlds, a bit from the product worlds, marketing, a bit from the growth worlds.

Eyal David: [05:49] Yeah, yeah, a perspective really, opening up and evolving, right, right, right. Totally. Okay, shall we continue? Yeah, we can, we have six more who answered, but we can scale. So okay, we can also move on, let's proceed. Oh, I don't know if we can engage the scale, okay. But fine, come on, so basically what do we see, how does it break down? Roles. So we have 61 product managers, we have two who are product owners, and 26 percent other. Okay, so that's a herd, we'll leave it, and basically 61 percent product managers, so we'll really tailor the conversation accordingly. Over to you, I'm closing the product. Yeah, I just want you to tell me, what's one or two things you most want to learn from this session? So the thing people most want is how to lead strategic, disruptive changes in product management. Alright, and how many know Fiverr? What, like, roughly one? 25 percent very well, 10 percent so-so, 50 percent... I've heard the name but have no idea what it is. Okay, let's give you... and another 15 percent not at all. Why describe it? Here, we can even move to the next slide of the "I want,"

Shai Shefrank: [07:01] right? The wish, the wish, sorry. So, Fiverr actually started... yeah, Fiverr is a company that belongs to the generation of companies after the bubble burst of 2008. All kinds of these quirky companies sprang up, Uber, and Airbnb... what's it called? And yeah, it was a kind of era where the world sort of discovered the internet, of course lots of user-generated content and things of that sort, and marketplaces, like, exploded, right? The... Airbnb, the... And Fiverr came and said, what's our take on this thing? I will do it for $5. Half a serious company, half a gag, right? This thing came out, and you could say, pretty simply, too good. Because, like, there are, I assume, dozens of podcasts with Shai Wininger and Micha Kaufman, and you can hear about how this thing was born. But broadly, I think, the concept, the format lit people up, and really quickly, there were thousands upon thousands of people from all over the world who said, this is what I'll do for $5. There was a moment, I think, they blew up on TV or something like that. Yeah. Got up in the morning, boom, the servers crashed, the pipe isn't working, and that's it. Someone dancing without a shirt, and someone else doing something. And really, at the beginning, it was a bit of a wacky place, right? Because it's a game, it's like, what do you do, are you crazy? What am I willing to do for $5, right? Fiverr, for $5. But this thing kept growing, and more and more people came to buy things that were legit, right? It started with logo designs and writing and translations and all kinds of things like that. And basically Fiverr gradually became a market, a market of services provided by — I won't say freelancers, right? Even though many of them weren't really even freelancers, to do gigs, I think they do all kinds of things, right? But there was product-market fit there, and this thing exploded, and I kind of came in at this stage, right? The most interesting company to be at, a consumer company in Israel that's exploding on the internet, it seemed to me like the biggest dream, like, you don't have to imagine how to do it. And then this company reached some point where it had to, or wanted to, change, and stop being this gag, and start being a marketplace for businesses that buy from freelancers. Because the truth is a lot of it was that. Right. And obviously this business is maybe less viral, but it's more business-like, it's more sustainable and bigger, and there's very significant potential here. So that's sort of the end of... You could also say, let's say how long they actually ran with the $5

Eyal David: [09:49] until there was actually a change, so let's say from... you sort of started 2010-ish, you actually led this in 2015-ish,

Shai Shefrank: [09:57] 5 years they worked. In 2014 — no, yeah, there was an evolution there, I think. There was an evolution, a marketplace, they added to it, there were all kinds of things called extras, every service was $5, but people could add on. There's a gig, and inside it there are extras, and the extras were completely custom, each seller could add whatever they wanted, no matter under what title. So I'll make you a logo now for $5, that's the title, but if you want the source file so you can also use it, then add another $4-5. Yeah, but in the same breath I could also add, hey, I'll also throw in, I don't know, I'll email you the file, or something totally unrelated — there was no productization whatsoever of the vertical I'm in, I could do it completely custom. It was custom, it was so custom, people wrote about what they charge extra for, and slowly what we actually saw happen in the marketplace was that more and more people asked for extras for the same things, and basically the whole marketplace was I will do something for 5 dollars, but in practice when you came to buy, you had to start assembling it there, right, adding on, I'll add another $20 for extra fast, and another $15 to get the source file, and so on and so forth for all kinds of weird inventions that people also came up with, and it was kind of something a bit messy, right, a bit — everything's five and also nothing's five.

Eyal David: [11:18] So we even have a screenshot of the homepage, I don't know if you can see it, it was all at five dollars, a whole mix of things like that, and there was also the nightmare, I remember, of doing the curation of this whole thing, but also keeping all the junk out,

Shai Shefrank: [11:34] right, every day about a thousand new gigs or something came in, a thousand sellers, I don't know if those are crazy numbers you quoted there. There were in the orders of thousands of gigs coming in every day, the truth is this was the first project I did when I got to Fiverr, they basically told me you'll do the back office, and it was kind of a bummer, to this day I think I almost didn't enjoy anything as much, like they still use it in some form to this day, right? Yeah, but like, I took it to a real product place, I said, okay, today they want to do review, I really worked on review, how do you review these gigs coming in, it's a sea of gigs coming in, and then they had this old-fashioned back office of tables they'd scroll through, and you had to hover over the icon to see the image, and over the title to read the title, and do a ton of hovers like that, and actions sitting inside dropdowns, and yeah, something there. And basically I built them a real system like the cards, that got a shortcut, deny, approve, deny, approve, deny, deny, deny, approve, approve — they sat, and I turned everyone's screen ninety degrees so they could see all the information. No, it was amazing, it was an amazing project, and I really enjoyed it, and it also became a lot of other things, like we realized you don't even need to, you don't need to check the gigs when they come in, you don't need to check the gigs when they're delivered, and then we created something that exists in Fiverr today, years after I left, suddenly, Made on Fiverr, that's exactly the deliveries feed, which we once did internally, for the editors, where they'd mark a diamond, and they'd pick the beautiful things that were really delivered in the marketplace. But it was really interesting.

Eyal David: [13:07] It was interesting, and it was a chaser project, Shaike. Let me bring us back into focus for a sec. Really, the situation that was created in the marketplace, in the catalog, was really unbearable also from the sellers' side, who basically had to adapt themselves. I'd say, a lot of sellers didn't come into the market, they went to Upwork and other competitors so they could charge more and provide more professional services. And also from the buyers' side — I was a buyer myself, I had to work, assemble a puzzle in order to get, what, a puzzle, meaning the relevant extras and... that's the know-how, I wasn't sure what I'd get within this whole thing.

Shai Shefrank: [13:38] Right, and that's really the place where, I think, as a product person, you meet both a very big business opportunity and a very big problem. And usually either we do opportunity hunting, or we kind of look to solve problems. But here there really was a problem that contained within it a very big opportunity, because this model was the core, it was the essence, right? And that was also the difficulty in changing it. The marketplace was called Fiverr, and that's a fiver. And you enter the marketplace and everything in it is at 5, like, it makes so much sense. But on the other hand, the seller can't sell the way they want, and the buyer can't buy the way they want, and there's so much friction. And I understood, like — not just me, and generally the company wanted it, it wanted to move the market up-market, it said, it's nice that there are all kinds of people in all kinds of third-world countries doing dances, and we want to sell serious things here, we have to move the market up, and it's also nice that people sell logos that they made in half a minute and look like garbage, to people who don't get it — we want quality people, we want quality deliverables, we need to take this marketplace up-market. But, we're called Fiverr. And it's not that we're unknown, it's not that this brand isn't worth something — like, we're one of the most famous gig-economy consumer companies in the world, at this stage it's exploding, what do you do. And there were a lot of worries, and I can say for example that before — before, there was a project that Likhai did, led it, of custom offers, right, where if you're chatting with the seller, they can send you a price quote, for any price. There were tons of worries, like, hide it, make it gray, make it small, make it off to the side, make it this. No matter how much we hid it, it made more money than the whole marketplace, because that's what people needed.

Eyal David: [15:19] So the signals said we have to do it, but reality said we can't. I'll say, like you say, there was a lot of fear, I really remember it, that all kinds of features of one sort or another that were going to break the $5, giving this freedom to the seller,

Shai Shefrank: [15:34] but they weren't follow-through, and it was clear we had to rip it off like a Band-Aid and make a much bigger move. Right, and then I think the move started, now, and here I am, it's a lot of people involved in these processes, right, me, me, and also ten years have passed, right, so I remember things the way I remember them, and surely everyone remembers things the way they remember them, I don't want to take credit from anyone, right? But some kind of conversation began about, so what will we be? What is Fiverr? Right, Fiverr was, I will do something for five dollars. And I remember the presentation I made — again, it's possible it was made before me too, but I remember saying, okay, I will do something for something. What does that mean? What's left? What's the core of the company now, when there's no longer I will do something for five dollars, but rather I will do something for something.

Eyal David: [16:24] Wait — I will, the "I will" is the thing we were left with, and this "I will" is a very very significant thing, because what does "I will" actually mean?

Shai Shefrank: [16:32] It means a seller, it means a freelancer takes a service and says in advance, here's what I'm going to do for a certain price, and this thing doesn't exist. No matter whether that price is five, ten, fifty — you want to work with a freelancer today? Go, write to them, do a brief, do a negotiation, decide on hours, days, this, this, this, this, this, and in the end somehow you'll manage to close on some price and you'll hire them. A nightmare. We don't want hiring, we want One click purchase. And so basically, it was very clear to me that the story had to change. The story is not what you can do for five dollars, but rather how you can buy a service the way you buy a product. The way we called it back then, service is a product. Yeah, so there was the whole "product is a service," right? We said no, service is a product. How can I, in One click — and I just read the...

Eyal David: [17:24] press release? The press release, so who's quoted there saying, it's like buying a flat TV online, right?

Shai Shefrank: [17:31] Ten years. Flat screen TV online. Yeah, ten years. It's like I asked my daughter, my daughter asked me what's a telephone? Why don't you call it — this is a telephone, this isn't a telephone. What do you want? So no, this is a telephone, right? Buying flat TV screen online.

Eyal David: [17:45] Tell me, and how was it? You like — you decided on... you decided you were going for it. You put your job on the line, first off, at max-gig. So I don't — I, what I remember is that I

Shai Shefrank: [17:54] was having this conversation I just described with anyone who would listen. But nobody stopped you, dude? No. No, it was clear to me that we had, we had, we had metrics. We need to raise spend, right? How much people spend in the marketplace, we need to raise quality. How do I raise, like, the satisfaction, the retention, right? In the end you want — and by the way, you also need to not hurt conversion, which is also a story, right? When everything's five dollars it's kind of... Especially since we were pre-fundraise, right, which is what was pressuring us at the start of the day. Yeah, we're a company, we're a company that has to prove everything, we want to show that we're growing, and all the metrics always need to go up. Yeah, nobody can raise... Yeah. And I, as a product person, understand that these are the metrics, and I do what I can to make it happen. And I do pitching. By the way, today at Empathy, for example, pitching is part of what product people do. We require them to prepare pitching. For their projects. What does it look like? Let's open a parenthesis for a sec. What does it look like? What do you mean? We don't work today at Empathy with any methodology of agile and scrum and all that. So... once a sprint. We work with a methodology similar to one called Shape Up, which is a methodology of six-week cycles, with a cool-down week between cycle and cycle. Where in each cycle, every product person comes with pitches that they bring, and they pitch them, like. They present them, and then basically we, the leadership, decide how we want to invest the company's money in the coming cycle. Which is actually time, development time. Yeah, it's money, it's development time. How do we actually, how many weeks are we willing to give each pitch?

Eyal David: [19:29] What happens if something doesn't come together, do you still want to do it? What does that mean? Something now takes six weeks, say, four weeks,

Shai Shefrank: [19:37] and you're willing to invest from your budget two weeks. So that's exactly what happens. So what, do they go, does he or she go? So we tell them, it's like saying, an entrepreneur comes now and says, I want to raise 5 million dollars now to grow this thing, and comes to an investor who says, cool, you've got two. And they've got two, what do they do? Either they drop the idea, they say, I can't, or they find a way to do it with two. We usually, you know, if we know there's no chance this thing will happen, we won't set it up to fail, but very often we challenge, we see, we see, cool, it's not — it's a nice idea and it's nice, but go find the version, right, in the Shape Up methodology, from the Basecamp book, there's this bit of like, find the version of this thing that can be built in a week. Not downgrade it. And they find it. So we'll expand this group, really interesting,

Eyal David: [20:27] I'm closing the parenthesis — you need to go back, so okay, you prepared the presentation, you pitched to the whole world and all,

Shai Shefrank: [20:33] now what? Now I — by the way, what is a presentation? A presentation is first of all starting to say that the company still has a story and the story is simply different. And the story is service is a product, and this thing has a strategy. Strategy means a lot more things I did afterward, that we're a marketplace where you can do advanced search, it has filters, it has all kinds of things it didn't have at this stage. I can say I want extra fast, and it'll update the prices of all the gigs, I can say I want a commercial license — we even did some small evolution of this thing there — but in the end it's about coming with real designs, like coming with these very well-defined wireframes of how this thing is going to be. In addition, we also did some kind of survey,

Eyal David: [21:17] that we sent to a lot of sellers, good sellers. Wait, let's stop here for a sec, let's stop here for a sec, you did a survey.

Shai Shefrank: [21:24] I do what's needed to prove that my claim is correct and valid and you can go with this thing. Great. Even though I'm shaking, and lucky for me that Asaf Porat was by my side to tell me no, let's do it, let's do it, and when he starts shaking then I want to do it. We're cute. Let's run some additional survey, I think fifteen who were with us are still with us, and basically, yeah, let's ask what you would have done in this situation.

Eyal David: [21:52] How would you have validated an idea like this? Is it a survey like Shaike's or something else? I'm putting up all kinds of options. While we wait — if you tell us when we can continue — there was actually within this story also something I learned,

Shai Shefrank: [22:05] it's important to understand that at Fiverr there are both buyers and sellers, and even if we decide as management to do XYZ, we're always dependent on the sellers actually agreeing to make this move,

Eyal David: [22:14] and that's really what you brought about here with the feasibility thing you did, or with the proof.

Shai Shefrank: [22:21] Tell me, how long did this survey last? Were the responses sharp, fast? It's very little, meaning one of the things that was fun about Fiverr was that there was scale. Meaning, it could take a sample, remember? Put out a survey and get 150 responses within half an hour easily. That's what you do, meaning, I'm telling you, I don't remember its exact details, but broadly we asked people whether they think it's a good idea or a bad idea, those who think it's a bad idea, we tried to understand why in follow-up questions, some open, some American-style, and those who said it's a good idea, then we tried to understand with them what it means for their business. So did you really look at this vertically at that time, or at all? No, I don't remember, I don't remember. But I think we sampled sellers whose opinion we were interested in hearing, we hand-picked sellers, who are sellers that work, sellers who have a service that works at five dollars, right, we wanted to. And what we discovered is that the vast majority of sellers think it's a great idea, and were also very very happy and very very able to tell us what would change the moment they could price the way they want. And for example it was important for us to know, like, what would be the new price of your service. Wow. Whether — because we wanted to anticipate, like, what would happen? So the good sellers of the five dollars, where would they be? Would they be at 200 dollars suddenly, would they be at 40 dollars, would they be at 30. First of all we saw that the element of worry and competition is preserved. Meaning, it's not that people say, I'm now, uh, I'll sell for these. Everyone always came and told us in the feedback, this is a highly competitive marketplace. Like, it's a price-sensitive marketplace. People come here because they want to buy for a fiver. Everyone was very very aware of the risk of raising prices. But many of them said it would save them a lot of back and forth. And it would save them a lot of the cancellations they make. Right, that they buy at five, but they actually don't want to deliver this thing, and... and it would save a lot of a lot of heartache. And to save this heartache, they're willing to play with the price and write 40 dollars on the gigs. And that really was the range of the prices, meaning, it's not that at this stage there were — after that we also did Fiverr Pro, and we did things, and this was upselling. That also pops when you can talk about them, like...

Eyal David: [24:25] Wait, but let's stay here. Let's stay here, one sec, so what came out of the survey? Just curious. So right now we got a majority of 60 percent. I would have tested the new pricing with a smaller segment first. So let's talk about this for a sec. This was true at Fiverr, which was divided into catalogs — tons and tons of types of categories, different types of selling, from different places in the world. How would you have approached this, since you shared about it after all and I'll show. My feeling is that it's actually not the way in a marketplace like Fiverr, for the reason that you have to...

Shai Shefrank: [24:57] There are things you can't roll back, okay? So it's a marketplace, in a marketplace first of all a seller is a community, it's not everyone operating alone, right? If I now decide that I'm going with 100 sellers and I give them something I don't give the rest, first of all I've created with them a mess at 90...

Eyal David: [25:14] Maybe tell about the guy who came to the offices? There was a seller who was... in New York, right, they made some change to the algorithm and he came to the New York office

Shai Shefrank: [25:22] and walked in and asked to talk to someone else about the algorithm. Lucky he's in Israel. No, but I have more examples, I don't know if you remember the change I made to the rating system. Of course I remember, I was more intense. Do you remember what happened? Of course. There was an internet strike, like, everyone closed their gigs and I thought — spark and sigma — and that same day I wanted to jump off the roof. You can ask Micha about it. To this day he remembers he had to give me cold water. No, so there are things — now, a feature of this kind that people want so much, you can't release it and take it back. People changed their pricing, you did this, meaning, it's also a change on the magnitude of an almost company pivot, like, business-wise, meaning, this company bought its name on being everything at five dollars. You can't now say, wait, I'm trying for a sec to not be who I am, but if it didn't suit me then I'll go back to being who I was. It's not the kind of change that, in terms of the messaging, in terms of the fear, in terms of the brand, in terms of the community, in terms of the ability to roll back, right? An experiment is something you want to be able to roll back from if it doesn't work. This is the kind of thing where once you touch it, you're committed.

Eyal David: [26:29] Let's stop, let's talk for a sec — so okay, that moment where you took a presentation now, did sampling, came back with data, this continued, now this was also with the accompaniment of leadership, right? Of Micha? With accompaniment all along, the whole time. How did it look from there? You actually came back with the data that says, yes, we want it, okay, so that's it, you do it, you launch a project like this, how long does it last, what does it look like, how does it go down? That I remember, you know, I remember the moment

Shai Shefrank: [26:54] I'm presenting this here, in this building, in a conference room upstairs, so there were a lot of conversations before that, but I remember the moment that Micha, and I think Dahlia was there — Dahlia was summoned, the company. Marketing. He was revenue, a CRO sort of, and I remember Micha, I don't remember whether he says it or I just understand it, but it's a specific moment, I want this. Let's go, do it, and I think by the way to this day — I've had a few readings with Micha already in my career, I think as a CEO, meaning, the amount of leadership and the depth of letting people lead significant things in the company, and going with them, like, and really going for it. Amazing, amazing. She's amazing, like, and I really had a home there, at Fiverr, to do things. The more I dared, the more I went wild, the more I found there was a higher receptivity, and I think with great wisdom.

Eyal David: [27:48] Great wisdom. Totally. It paid off. Okay, so they decide to go for it. Decide to go for it. What did this project look like, because you now have to make the s... beyond communicating it, beyond making this whole project happen, and also understanding what it means. There's now a row of sellers who now need to change their gigs, and it won't appear at all in the live site, right?

Shai Shefrank: [28:08] How was it actually? So — what, there were, there were big technical questions here, like, of how you keep being, after all you have to support, and you have their marketplace, if — I don't know, there was burden on it surely — tens of thousands of gigs, many tens of thousands of gigs, that are active and selling, I'm not talking about the long tail, which is maybe another hundreds of thousands of gigs. And there's no end of gigs, and they all need to work, right? Now, you're changing pricing, which is not a thing... it's at the core of the core of this whole system. First of all, we decided, in terms of pricing, in order to preserve the — so to speak, service as a product, that would preserve the ability to buy in one click, we basically introduced something that's sort of a kind of packages. Meaning, we'd let the seller create their gig in minimal, standard and premium offerings. And each would include different things, like for example, yeah, there's a design to it. And this still seems to me, to this day, the kind of thing you see on the gig page at Fiverr, if I'm not mistaken, and even though I bought there, but not really. Yeah, yeah, we're pulling it up now, yeah. So we decided to go with this page, it was also very very hot at that time, because all the folks were doing service as a product, the whole internet was full of these kinds of tables, and how you can get people to buy this way or buy that way, and we also fell in love with this conception, this concept — not conception, it's a concept, and we said, this is how we want to do it, and it was also really right for us, because we wanted to let the seller say, here's a small project, here's a medium project, and here's a complex project, like, a big project, but with that, to be done — because come on, think a bit, say what's your small one, what's your medium one, what's your big one, and then basically what we did was that all the old gigs became kind of a single package, right. They have one package, and all the gigs that made the transition became kind of a triple package, meaning, they had — and then that allowed us to keep supporting the backward experience, and also still allow, and it's also fine, they said, I don't want the thing, and I'll stay with one, it stayed one. There was a lot here, there was crazy work here in the world of categories, which afterward became category development, which today is a whole domain within Fiverr, of how I price different categories, because obviously in order to price design, I need to allow myself to add on for the source file, and for a commercial license, and for the number of characters in the illustration, and for whether it's colored or not colored, and whether it's used in stock or all kinds of things like that, you need to add money for that, so what we did, we did a very big project, of going through category by category, even sub-category by sub-category, and defining what the pricing factors are, that's what we called it, pricing factors, of each of the categories. And according to that, really doing this build. So it's a big project, it's a project that lasted, I think, a quarter, if not two, like, in terms of engineering, very very technically complex, but really the day after it was released, a new marketplace was born,

Eyal David: [31:00] everything changed. Tell me, totally, tell me, how much, one, how complex was it for the sellers? Because suddenly they have to price this, I don't know, we put a screenshot here of a category of, simply, probably it's translation or something like that, of a proofread. Suddenly they have to price it per word. Like, how simple was this transition actually?

Shai Shefrank: [31:21] Until now, whoever translates, they priced per word even before. They just had no way to do it. So you went up and actually looked at how they work today, just went up on them in this specific case, and then you came out with this calculation. Totally, we created, we went to each category, we figured out what are the extras they perform, that all the sellers do in each category. Here, in translation it's number of words, and it's delivery time. And it's sometimes, I don't know, maybe there could be another proofread, some other thing, I don't know. They found all kinds of ways, right? So we took these things, that most sellers did manually, and we made them structured, simply. Once we made them structured, it also became a search filter. Now I can enter the marketplace, say, I want a translation of 500 words, from French to Russian, in something, in three days, and already all the prices I see in the marketplace update according to the prices of the services. What, it's amazing, that this is something that came up bottom-up, right?

Eyal David: [32:17] It's a move you wouldn't expect to happen that way, in any form. You'd expect that just, someone would first come and say, we're doing this now. I don't think so, because I think this is a move that's so so embedded within the core of the product, that leadership isn't there. Leadership — and good thing. But I don't expect a CEO of a company to now know all the extras people do and to understand at all that in different categories people price differently, and this and this and this, and to manage to understand that there's some product picture here that solves this thing, here there is — for a thing like this, you need an entrepreneur inside the organization. But you can't expect the CEO, who's also an entrepreneur, to take this thing on, because here there's something that really, really has to be — you have to be inside the market of the... of the problems and the connections and the ins and the breaks, in order to understand that there's something here, and also I think it's such a project, that already generates worries, so many things, so many, that you can't drop it on someone and have them do it properly. It's also — so it could be that it's partly a matter of luck, if you have this person, right, these people in the organization, if you have this culture in the organization, that allows people to take such things and do them. Tell me, and how do you look at this now? Like, in terms of the fear that you felt, that those around you felt about floating such a thing, is it something you've gotten to do again since? You went and now went against the current and pushed such an initiative inside the company?

Shai Shefrank: [33:46] Do you see it at Empathy, in the kind of interest? First of all, I don't think it was against the current. I think it was, again, I had very very good backing there, and again I bring back the disclaimer from before, it could be that Asaf would listen to the podcast and say, what are you talking about, that was me, yeah, and it could also be that he's right. I, everyone — if there's something I learned at Bezalel by the way, in design studies, it's that ideas, everyone in the class, I think, thought about the same one, in the end they come to the submission, everyone submits the same thing and each one's sure everyone stole it, because ideas just float around, and sometimes you tell someone something, they hear this thing but they're not where you were connected. Even 10 years... no, and then you fix it in your head and only after two months, you even understand what he told you and then you say, wow, what a crazy idea. A friend two months ago told you this idea, but you weren't in a place where you understood at all. So I say, I'm putting a parenthesis here, I also don't... you know, I kind of have... but I don't remember what you asked.

Eyal David: [34:44] I'm going to come back to it, Mia, do we have more questions piling up there? Because this is the time to ask questions, if you have any — somehow, this really was a crazy move. Fiverr also rose from there like crazy, like they started, suddenly categories opened up, sub-categories opened up,

Shai Shefrank: [35:00] lots, lots — we'll do for the company and... you mostly encountered us during Covid, in the crazy surge the stock had. Yeah, I think Covid is already a good few years after, yeah, it's already four, five years after. There's no doubt it changed everything. Meaning, it enabled a ton of people who couldn't and didn't want to now deal with... so much, people who say, I don't put a price tag of five dollars on anything I do. Because that's exactly why I'm not okay. And on the other hand also the marketers, I don't buy any service for my production at five dollars. Like, there were a lot of people for whom this amazing positioning that Fiverr went through, it simply distanced it from that world and that solved it. Meaning, pretty much immediately we started seeing more serious people coming in. Obviously everything related to ASP, right, to the raise, went up very fast, but yeah, so it also raised a ton of other challenges. Yeah, meaning, also, let's say a person who comes to buy, they have certain expectations, something five dollars, also after, when they buy the same thing for fifty dollars, the reel of their expectations is already completely different. And suddenly it doesn't work. In the past they'd buy, they'd say, okay, it's not five dollars, I'll buy ten, don't feel like it and I'll toss it, keep it, like. And this created, first, it created a lot of dissonance and it also created a dynamic and expectations suddenly. So these things always come with difficulties too, right,

Eyal David: [36:26] it's this thing that in a lot — of a lot of software, really was its essence. Shai, this is a lovely story, wonderful. I want you to stay with me for a moment at that same moment. So basically we talked, again, it's a hero's journey, right? You went out here with an initiative on your shoulder. Basically what you're saying is, first of all, to open up all the... your product people. That's additional, you searched for the problem — we also have a question, I still want to... okay? And then we'll stay with the question. More questions, if there are any, again, that we'll do briefly — so if we look for a sec, so this move was first of all to identify the opportunity and the problem, there was also a crazy opportunity here, which is amazing, you went and actually shared it, and actually built the story around it too. I also understood what the main blocker is. The main blocker was that the company is called Fiverr, and I want to stop that. Every... in the company the... blocker, not a small one, it's fear, worry. By the way, worry — we didn't get to talk about it, right, but it was maybe a core topic. I think worries, I see it with product managers I work with today, and I saw it with product managers I worked with on many occasions, and also with founders, and also with... all of us, with human beings, and I'm not free of it, worries, fears, it's a very very strong force. It's your engine, but it drives you.

Shai Shefrank: [37:39] That's what you're saying. It doesn't drive me, but I had to overcome it. You know, I can say today at Empathy. Yeah, we make a product that's meant to support people who are dealing with the loss of someone close. How many worries there are there. How do we say, how do we do it, how do we push, send a push, don't send a push, do we say good job, don't say good job. It could be that we need to free ourselves from all the worries, that's how we make the product better, more impressive, clearer, faster, fairer. Now, our worries are ours. The world, it doesn't care that you have a past at Fiverr. It doesn't care, it wants to buy something nice now, at a nice price, without now racking their brains and doing, assembling and something. Fine, they want to sell their service. It's very nice that there's Fiverr for that, and it really brought success, but fine, it's very nice that a person now works for someone, but they need now to close an account with a cable company, and let's help them do it, like that's it, so we'll apologize already.

Eyal David: [38:39] That's it. It was here, but not — it already became fear. The moment you went to... at Empathy, and you already put the pitch of it already out there, and you, they expected from you, you could no longer go back. They also expected from you the staff, and all the weekly meetings you had,

Shai Shefrank: [38:53] I don't remember, but was there something like that? A couple of days, so I remember, I think it drove in a... a broken apartment that didn't...

Eyal David: [39:00] But that's exactly the entrepreneurial thing you did. Right, you put yourself at the edge, and you couldn't go back from it. So that's actually the thing I took from you, something very right. You also told the story very well, you also talked about how you actually hide the screens, in order to talk to them... here there are also very visual people, that you need to really illustrate for them, not just with mockups, but really designs of what you illustrated, how this thing looks. And you also neutralized risks, and you know, we always talk about this. You went and actually described this survey, described bringing people from all kinds of categories also afterward and understanding more... already, so that it's easier for me to understand that I'm not making a mistake. That's it, and it's very important. You talk about fear, and I see product managers, whose fear stems exactly from there, because there are a lot of holes in the story. But don't let fear stop you, ask yourself how do I bridge over the fear,

Shai Shefrank: [39:43] like, another good Fiverr story, I'll tell it in half a second, if I may. After we stopped the rating system — at Fiverr there was Thumbs Up, Thumbs Down. You gave the service, you received it, now how was it? Thumbs Up, Thumbs Down. What happened? People, either if they were really pleased, or if they were okay, they did Thumbs Up, or if it was really crap work, they did Thumbs Down, because like it's unpleasant to boo, like, to shame the person who... And then the whole marketplace was, everything at five stars, a hundred percent rating, and everyone's pleased, everything's at 5 dollars and everyone's pleased. So that's exactly it, the matter of stars was, there was Thumbs Down, but it translated afterward into what your rating percentage is, and how many stars that means. And then I said, I want to move to a method, right, and also. So also Thumbs Up, Thumbs Down, Thumbs Down, Thumbs Up, Thumbs Down was a way to rate things, for a lot of reasons. And I said, I want to move to Star Rating, and I want to ask all kinds of questions about all kinds of categories, and that's what makes sense to do. And then, also, you know, founders came, so I think it was also with Micha, and said, but wait, today it's all max... well, because everything costs 5 dollars, and everything has 400 reviews and 5 stars, everything's perfect. Like, what will happen? People really, we know they sell, it's just cheap services, like what, you're going to switch, what will happen? My whole marketplace is going to be three stars, like, so I was forced to — so there's fear here, so I said, wait, how am I going to answer the...? This is a justified risk. So what did I do? I built on top of the Thumbs Down, Thumbs Up system, also a star method, and I tagged it as internal rating, experiment, and I asked the questions, and then, no, I wrote experiment, really, and I collected over two months on every order at 5, what its Star Rating is. And then I looked at how the whole marketplace looks in Thumbs Up, Thumbs Down, how the gigs look, which is everything either at 0 or at 100, very much at 0, and all the rest at 100, and how it looks when... when in Star Rating, when I still showed 80% of the marketplace keeping at 100, and that's the rest, and then basically I was able to tell them, listen, here, an experiment ran for two months, this is how the distribution of gig ratings in the marketplace looks, after the change. This is how it looks before at all, this is how it'll look after we make the change, nothing to fear. And only then, the founders told me, okay, if that's how it'll look, then let's go for it. I went for it, and then the sellers freaked out, and said, what is this crap? I'm closing my gig, this is crap. They called it the Advanced Rating System, ARS, right, which is ARS, like, ARS. They called it The ARS System, like, and there were some five hundred comments on the blog where we announced it, where everyone said, this is the worst idea in the world, it's going to ruin the marketplace, I'm closing my gig, never, it's terrible. In the end everyone put their gigs back up, and everything stayed fine, really the rating, like, improved significantly, in terms of representing quality, slightly. But also, it's not that I say, not scary, but I still do it. It's saying, wait, what's the risk, how do I bridge it? Like, so I'll do a survey, so I'll do this, so I'll understand what they want, I'll do, understand what will happen when I make this change in the future, so that I can, so that I can speak to one thing after another.

Eyal David: [42:51] So, to isolate for a sec, really the risk, let's dwell on this for a moment, it's also something that happens in the experience at Fiverr after the... when it's done, the buyer already receives the deliverable.

Shai Shefrank: [42:59] There's really no big risk here of them not converting to another gig, or to the... the rating? No, the danger to the deliverable is that if all the gigs,

Eyal David: [43:08] instead of being five stars become three and a half, then there won't be conversion. In the effort you made over these two months, that you actually added some piece to this puzzle of Fiverr,

Shai Shefrank: [43:16] that's not supposed to affect, or like at the risk level, whether for a sec it's supposed to affect. Right, first of all sellers didn't even know it existed, because only owners saw it. And... yeah. And it happens after the order, and yeah, it was completely internal.

Eyal David: [43:29] But it's not something an ordinary product person would think of, I think. It's a very entrepreneurial mind at the end of the day.

Shai Shefrank: [43:37] I don't know, maybe. But I think that with an ordinary product person, and I'm sure, right, I always advise a lot of product people, and there's always a full product tendency. Right. Why don't you do this, why don't you do this, why don't you do this, do it, what are you doing, do it. That you don't have to do the move... No, but management, but they're afraid, and that's... okay, find a way to bridge over the fear. Like, you're afraid? That's not my problem, that's your problem that you're afraid, you can't do your job. So how do you now manage to do some proof of concept, or some mockup, or some experiment, or some survey, or some thing, that helps bridge over this fear, because the fears are justified. Or how do you manage to understand that if a story has run its course here, then you need to help create a new story. I'll say one more thing... it's more moving, you have to move on from this story. And what's also always worth looking at is what works at the competitors'. In this case, you — you weren't the only marketplace at that time, there were good marketplaces that did the rating thing well, you can look at what they do, there you go. And also... we'll get great ideas, and it's also the standard in the industry here. That's what you do.

Eyal David: [44:37] Wonderful. Do we have anything else in the summary we didn't go over, I think we summed up not bad, we even jumped into additional parentheses. If you have anything else, then you can... we also have a question, right? There's more than a question? Wow! We have 3 questions. Wonderful. An anonymous question, how do you price in the industry for different things, for example, pricing per word,

Shai Shefrank: [44:59] but design per project, complexity? So here it's a process. First of all, in order to solve this thing, in the end there arose at Fiverr this whole arm of category development, which is something I did, I did it there, really one of the most complex projects we did was the calculator for number of words. But it really was a product we developed, for categories, that are categories that need quantity as a... like, it's voice over, it's writing, it's translation, it's all these things that you price per word. But that was kind of the thing. Meaning, at the beginning we started with things that are more freeform like that, and the more we wanted to do more structure and more standardization so that people could come and say, I want 500 words and have it compare the prices for them across all the gigs, we had to develop real pricing engines, per category. And then how do you do it? You start from the big categories that generate the most money, that have the most... prioritize them to the top. Prioritize them to the top, also those that have, say, complexity, right, say, things that are just to say now, I'll give you or won't give you the source file, I'll give you or won't give you a commercial license, I'll give you or won't give you — these are easy things. It's boolean. So first you create the booleans. After that you start to say, okay, let's do those that are not yes or no, but rather how many. Right? How many? One, two, three, four. Yeah, that's easy. After that you move already to words, which is also quantitative, but it already starts to be in much higher multiples. And also with words, suddenly you understand, wait, between translating words it's also related to delivery. Meaning, it could be that I can translate fifty words a day, but five thousand words I can't do in a day. And then you start to say, okay, so in these categories I start to create relationships between delivery and them, and every time you release such a thing, the sellers say, well, come on... wow, you solved, you solved a problem for me, right? So you kind of try to move between the different categories, where there's potential, where there's a big problem, and you say, now I'm investing two weeks to solve this pricing point. In this category, for these sellers, doing, talking with them, because I send them messages, tell them, if you now have this thing, the adoption is always very very fast, like, the strong sellers, it's their job, they immediately pounce and take it.

Eyal David: [47:21] There's another excellent answer here, there was also a matter here, probably at some specific moment, you identified potential, and in some category, there aren't enough good sellers,

Shai Shefrank: [47:28] so suddenly you also need to actually handle it and bring in sellers, that's also not an easy move. That wasn't really a problem, meaning, supply was never a problem for Fiverr, of good sellers, but notice, I'm saying, you now have folks working at five dollars, suddenly I add here elements of pricing, that take them out of their comfort zone, that now require better sellers, so where is that in the prioritization? I think one of the things that happened at Fiverr, the core of Fiverr was in the worlds of affordable services, so there was enough commodity in the marketplace, meaning, and the moment we opened up, and removed the five dollars, they came, they came and we also saw them, meaning, suddenly you see someone selling a logo for 200 dollars, versus tons and tons still selling for 5, 10, 15, 20, meaning, wait, who's this? Suddenly the editor, we'd do all kinds of pings for them, here, someone sold for 400 dollars, look, suddenly we'd look, you say, wow, this is a German designer who studied at the Hochschule, he makes beautiful things, come on, let's do a promotion for him, that was the easy part in the organic worlds. There was another move we can talk about in an episode too, which was Fiverr Pro, that's what I did from New York, which already means how we become a marketplace where you sell a logo for 5,000 dollars, for 50,000 dollars. It's a very interesting move of a strategic up-market leap, that there I already did, it was also something I drove and led here, look, we brought the designer who designed Apple's logo, to logo and design at Fiverr, so that they'd... like, we brought, there were stars, contributors, like, I remind you of bodies. There were really celebs who came and sold on Fiverr, just to say, here, Fiverr can also be ten thousand,

Eyal David: [49:11] find, you have to do it in the middle. Okay, we'd really love to hear this, but we have two more questions, right? A question comes from me, how has your thinking in product development at Empathy changed in the new world, that combines the new AI tools? The new AI tools. I want to ask the question about Fiverr.

Shai Shefrank: [49:29] I no longer think about Fiverr, I think a ton about Empathy, Empathy at its essence, a service that combines, a kind of self-guidance, but also a human service, we also have a care manager. It's very very important to us to keep our service human, that our service is human, in the places where it's important that it be human, but it's also very important to us to allow our service to grow and scale. I think AI for us is mainly a tool we use internally. Meaning, our care managers can use AI to shorten their work for themselves, so if for example in the past a care manager, after a meeting, had to summarize the meeting, and enter it so that in the next meeting they'd know, what they talked about with a certain family, then today, for example, we have an engine, that knows how to shorten the meeting for them, and knows how to write the tasks for them, and knows how to generate tasks from this thing, and knows how to extract the... a lot of efficiency. A lot a lot of efficiency in order to actually enable one care manager, for 20 families, for 200 families, and for 2,000 families down the line, but it's still very very important to us to keep this service a human service, in the places where there's the human interface, and I think that, bottom line, in these worlds of loss, but in the places where it's possible, in the places where, I'm looking for information, and I can find information and it's reliable, and it solves the question with the user, etc. etc. There's a lot of thinking about, about improving the service.

Eyal David: [50:59] Nice. I think that's it from me, so let's move to the last question. Yeah, one more question from me. What's the biggest change that's happened in the way you manage product today, with the entry of Generative AI?

Shai Shefrank: [51:11] First of all, I waste a ton of time trying all these things, because I play with them. I think also here, I think mainly mainly, it's related to my personal productivity, like, my ability to simply raise my output. To improve my communication, much faster, we also work in worlds that are B2B worlds, I communicate a lot with partners, communicate a lot with Americans, corporate America, from the world of Life Insurance, for me, this thing is super powerful, meaning, for me to write an email to a VP at an American financial company, in the past I had to work harder, much harder. So I think that there it's without a doubt, totally totally, I think a whole lot, like, we use them, by the way let's say, in the audio worlds, we use them inside the product, in that, transcriptions, things related to accessibility, quite a few tools, I think, also in terms of, in terms of my personal productivity, it's nothing less than a revolution, and I think most people are grasping it, simply because it's at your fingertips, you can do a ton of things with it.

Eyal David: [52:29] So give us some use case, okay, look, now there's a thing for emails, and things in that style, proofreading, are there additional things you do with them? Tons.

Shai Shefrank: [52:41] I'm roughly, an hour a day, I sit in Fountain — pound — boss, and talk with, and talk with ChatGPT, but you know, it's a kind of quick writing, take a provider and turn the... into a copy cause,

Eyal David: [52:53] I very very much enjoy working. Do you work with it in Notion too, a bit, right?

Shai Shefrank: [52:59] It does the tasks, right? All of Empathy's, like, all our internal documents, it's in Notion, so that sales can come today and ask, how does the obituary writing feature work, and get an answer, in text, that they can send via copy-paste to the client, right? It's very very powerful. I don't think — it's hard for me to answer this question, because it touches on everything in all my work domains, right? You know, actually if you were an entrepreneur, then an entrepreneur kind of at early stage,

Eyal David: [53:29] and at the start of the road, actually here you'd have a ton a ton a ton of value, because you could actually do a lot of the development work with the help of — I really see that a lot of people who suddenly have, you know, the one that does their marketing for them, the AI that does their marketing for them, their messaging, and then suddenly someone who does, someone, an AI that does their design for them, and then the design he brings down already to front-end, and he also has the back-end, which he does in no-code, and boom, he kind of created a product you can even charge money for, the marketing you also manage with AI, right, maybe in Gherkin-la it... an order does it, I think this is an amazing period, and there will be a much much more amazing period for makers like these, independent ones, right

Shai Shefrank: [54:11] there's no limit, I think this is a crazy game changer, but also in our worlds, that you can do a ton a ton of things with it and it's a crazy superpower, it's a gift we received, I look at it as an infinite army of workers half capable or capable of some things and I manage to get quite a bit out of it, wonderful, we have

Eyal David: [54:35] more questions, right? Come on Shaike, what a joy that you came. I had a lot of fun, we opened on, a lot of threads for more meetings like these, but not about Fiverr, enough, what, next time we'll do an episode for them, thank you so much to everyone who joined, we hope you got value, and we'll meet next time, bye bye