April 21, 2022

How to Launch a Marketplace | Ray Reddy, Founder of Ritual

Launching is often thought of as a huge event. Lots of noise, lots of PR. One big bang. For Ray, it's just a series of many small experiments. Ray realized that the number one value prop for a consumer was coverage: when a user opened the app, the majority of merchants that were walking distance needed to show up. So, he set up an experiment. He targeted a handful of office buildings and a dozen merchants, all located on the same block. He didn't need to onboard all the merchants in To...

Launching is often thought of as a huge event. Lots of noise, lots of PR. One big bang. For Ray, it's just a series of many small experiments.

Ray realized that the number one value prop for a consumer was coverage: when a user opened the app, the majority of merchants that were walking distance needed to show up. So, he set up an experiment. He targeted a handful of office buildings and a dozen merchants, all located on the same block. He didn't need to onboard all the merchants in Toronto, just all the merchants on this one block.

If you're about to launch a consumer product or a marketplace, check out what Ray has to say.

Send me a message to let me know what you think!

00:21 - How to Go To Market

02:25 - Convince yourself your idea sucks

04:38 - If you must, make it free

06:20 - Setting clear goals

08:08 - Finding PMF post-COVID

11:17 - Finding PMF gets harder the bigger you are

13:35 - Recap

WEBVTT

00:00:00.270 --> 00:00:02.609
This is the bonus episode with Ray from Ritual.

00:00:02.850 --> 00:00:05.400
Here, we talk about how he launched his marketplace.

00:00:05.700 --> 00:00:06.179
Check it out.

00:00:06.360 --> 00:00:11.429
Welcome to the Product Market Fit Show, brought to you by Mistral, a seed stage firm based in Canada.

00:00:12.119 --> 00:00:13.830
I'm Pablo, I'm a founder turned VC.

00:00:14.339 --> 00:00:17.670
My goal is to help early-stage founders like you find product-market fit.

00:00:21.690 --> 00:00:29.760
One of the things I'd like to dive on a little bit is when I hear the story of Ritual, I heard it many times, what always struck me is just the amount of thought that went into go to market.

00:00:30.030 --> 00:00:35.700
And you've described here the thesis you had, and you described, you kind of set up the experiment, and so on.

00:00:35.701 --> 00:00:45.929
That's one thing I see missing a lot, is idea side, and then maybe a scaling side, and this entry point piece of how you're really going to go to market is glossed over.

00:00:45.719 --> 00:00:52.229
It's like,"well, we'll put up ads or whatever." What led you to go so deep, and have so much thought on go to market?

00:00:52.591 --> 00:00:58.020
How do you almost...What's your advice, especially the first time founders, on how they should think about entry point?

00:00:58.229 --> 00:00:59.250
Just the philosophy around it.

00:00:59.579 --> 00:01:02.679
I guess my main advice is to think about it.

00:01:02.679 --> 00:01:10.560
I think that t's easy to not think about it, and to get very excited about things, but I'll give you an example.

00:01:10.590 --> 00:01:16.290
Building a simple spreadsheet model is very helpful, for most people, I think the minimum amount of work you should do.

00:01:16.290 --> 00:01:19.950
I'll give you an example of what I mean when I say the minimum.

00:01:19.950 --> 00:01:28.739
Marketplace, where you have two customers, for example, in our case, you should be able to reasonably answer the question"What is the cost of acquiring a merchant?" Do that now.

00:01:28.740 --> 00:01:31.799
Are you going to hire a sales person?

00:01:32.159 --> 00:01:34.140
What do you think the productivity of the sales person is going to be?

00:01:34.349 --> 00:01:50.640
And often you'll be shocked at the results, and you build that, and some of your basic assumptions on,"we may only need a handful of sales people." Then you're like,"oh no." If you apply any reasonable sales conversion rates, it turns out that you're egregiously wrong on some things.

00:01:50.551 --> 00:01:56.909
And when you say ads, the moment you take one, it's actually that ads is a bad strategy, but just take it one level deeper.

00:01:56.969 --> 00:01:57.840
Here's how I think about it.

00:01:58.379 --> 00:02:00.480
Building products is very difficult.

00:02:00.810 --> 00:02:02.129
It requires engineering.

00:02:02.430 --> 00:02:04.920
You are committed once you go down that path.

00:02:05.010 --> 00:02:10.680
My advice on this point would be, the thing that is free to do is to run business model scenarios.

00:02:11.129 --> 00:02:19.860
If we, in the sense of, you change variables on a spreadsheet, you don't have to deal with customer support, and ad creators, and brand, you don't do any of that.

00:02:19.889 --> 00:02:24.449
You just literally change numbers, and you can get a sense of how the business would work.

00:02:25.020 --> 00:02:26.550
For us, we built out a model.

00:02:26.729 --> 00:02:35.430
We said,"what would the operating model of this company look like?" I guess my advice would be, build a model to try to convince you not to do this.

00:02:35.490 --> 00:02:39.419
Because otherwise you have only happy data in your model.

00:02:39.420 --> 00:02:41.219
Everything is more optimistic.

00:02:41.550 --> 00:02:52.770
And I would say,"what's the point of doing that?" The way that we viewed it is, let's save ourselves from an immense amount of pain and torture if we can prove upfront that this is not a good idea.

00:02:52.770 --> 00:02:52.770
So, I think that most people go into it saying,"I'm going to do this to further convince myself that this is a good idea." I guess my advice would be, try to talk yourself off using a model.

00:02:52.770 --> 00:02:52.770
Because by the way, there's opportunity costs to pursuing an idea.

00:03:07.900 --> 00:03:13.120
It's the infinite number of other things you could do that you're not doing.

00:03:13.031 --> 00:03:16.870
So, it was very helpful for us to do that.

00:03:16.960 --> 00:03:25.270
And shockingly, many of the variables ended up being close to the few ethical values we made up, which was shocking.

00:03:25.270 --> 00:03:26.830
Things don't tend to work out like that.

00:03:26.830 --> 00:03:40.150
But, we made assumptions for example, on how often would an active user buy with us, what is a reasonable margin you could expect, what would it cost to acquire a customer, think the payback of that would be, is that even reasonable.

00:03:40.250 --> 00:03:50.349
It kind of forced us, even before we built the business, we had a rough sense of how would our team be organized in a city, what they need to do, how many people would we have.

00:03:50.349 --> 00:04:00.340
And you just start to think about,"do you actually need an operations person?" and like"Oh, okay, that actually fundamentally changes the P&L of a city if you need to do it.

00:04:00.341 --> 00:04:02.199
Is there an option for us to not do it?

00:04:02.590 --> 00:04:07.780
And to have, blah, blah, blah." And so, there's just a lot of...

00:04:07.781 --> 00:04:11.289
you can think to go to market in quite a bit of detail.

00:04:11.290 --> 00:04:15.849
The thing you can't test before you build a product is consumer reception.

00:04:15.850 --> 00:04:23.139
There are certain things you can do to get a signal around it, but it's pretty hard to actually test it without actually doing something.

00:04:23.141 --> 00:04:28.959
I think the go-to-market stuff is very easy to work through CAC/LTV, payback models.

00:04:28.961 --> 00:04:32.139
There's so much data online on funnel conversion of ads spend.

00:04:32.141 --> 00:04:38.319
And I think for someone who respects their time, it is a very prudent thing to do.

00:04:38.860 --> 00:04:39.009
Perfect.

00:04:38.860 --> 00:04:44.500
So, back to the story, maybe my first question is, if you think about that office building, how did you convince those first...

00:04:44.500 --> 00:04:46.029
because you have that chicken, and egg problem.

00:04:46.360 --> 00:04:49.959
How did you convince those first 15 merchants to sign up and deal with you guys?

00:04:50.769 --> 00:04:51.370
Just make it free.

00:04:51.370 --> 00:04:54.459
There are ways of doing it.

00:04:54.519 --> 00:05:01.240
And so, we saw it as,"let's go get these merchants no matter what." You can do this with no money at all.

00:05:02.651 --> 00:05:06.220
But sometimes spending a little bit of money, your time is also money.

00:05:06.250 --> 00:05:13.779
If you try, it's possible to do it with no capital, but that sometimes, in a strange way, that is actually more wasteful resources.

00:05:14.831 --> 00:05:21.069
For example, trying to convince people to try something may take a long time, but if you just say,"can I buy you a free lunch?

00:05:21.071 --> 00:05:24.370
" you tend to have a lot of people interested in trying it out.

00:05:25.240 --> 00:05:31.629
So, spending$10 on a$3,000 investment to sign up an entire building is not a stupid idea.

00:05:31.630 --> 00:05:33.699
Compared to, how else do you plan on recruiting 200 people when you have an app with no name?

00:05:34.319 --> 00:05:38.139
And I think the same thing is true with merchants.

00:05:38.141 --> 00:05:44.740
At the end of the day, even if you have to forget about charging them something, even if you had to pay them a small amount to participate.

00:05:44.411 --> 00:05:57.589
The way we saw it was the faster we can prove that this works, there's a lot of value because once we prove that it works, we had the conviction to invest for the long term.

00:05:57.711 --> 00:06:09.180
So, after the first 60 days we got out of prototype product-market fit questions mode, and we're like,"no, this was going to work." We convinced ourselves.

00:06:10.579 --> 00:06:13.009
The most important people to convince is not your investors, it's yourself.

00:06:14.389 --> 00:06:19.699
And so, we convinced ourselves that that data was sufficient for us to believe that we were onto something big.

00:06:20.000 --> 00:06:26.750
And did you have clear goals going into it, that if you get this engagement or this many of repeat users or whatever?

00:06:26.750 --> 00:06:27.800
How did you think about that?

00:06:28.430 --> 00:06:34.259
It's hard to set clear goals for a product that doesn't exist.

00:06:34.339 --> 00:06:38.420
I think that everyone, you have a pretty reasonable sense of what's good and bad.

00:06:38.661 --> 00:06:44.720
If a bunch of people try it, and don't use it again, that answers your question.

00:06:44.721 --> 00:06:47.060
If a reasonable number of people use it again...

00:06:47.060 --> 00:06:52.399
And then one nice thing, by the way, of doing it the way we did was, these people were so close to us, we could talk.

00:06:52.401 --> 00:07:01.639
And so, we had a beta group of people close to us that just gave us a ton of feedback, they were engaged, and the data spoke for itself.

00:07:01.641 --> 00:07:08.660
When I said,"when you get it right, the first time that is more luck than skill," that's what happened to us.

00:07:08.661 --> 00:07:10.879
I think it was luck.

00:07:10.910 --> 00:07:12.800
We thought of the century point strategy.

00:07:14.029 --> 00:07:20.000
If this didn't work, we'd probably have come up with a few others and tried a number of them.

00:07:20.180 --> 00:07:27.350
But this idea that we had that coverage was what really mattered, we should just really focus on that, and we have to get that right above everything else.

00:07:27.500 --> 00:07:35.509
And I still say this today when people join the company, no single app feature is worth more than giving a user all of the merchants.

00:07:35.839 --> 00:07:38.629
That is the number one thing we can do to give value to users.

00:07:38.839 --> 00:07:50.300
Basically, it's the same idea of Amazon's increasing assortment, that everything else is an accelerant, but that is almost a foundation of the more you have, the more useful it becomes.

00:07:51.079 --> 00:07:54.170
Yes, I think we proved that very quickly.

00:07:54.589 --> 00:07:57.259
And then we started to take a long-term view on things.

00:07:57.500 --> 00:08:03.139
Then we weren't building for a three-month experiment because nope, we're committed, we're going deep on it.

00:08:03.350 --> 00:08:07.579
And then we started to take a multi-year view on how we built from that place.

00:08:08.240 --> 00:08:08.750
Perfect.

00:08:08.240 --> 00:08:18.259
So, that's Ritual, and number one, you do this test, and it was one building, 15 merchants, 300 employees, it goes well.

00:08:18.261 --> 00:08:25.250
And then you start going neighbourhood by neighbourhood, merchant by merchant, always worried about coverage, expand.

00:08:25.250 --> 00:08:27.319
I think you were at 40 some cities.

00:08:27.321 --> 00:08:33.289
And then if we fast-forward, many years until COVID starts, everything kind of flips around.

00:08:32.961 --> 00:08:39.860
All of a sudden, you're focused on because you're thinking about coverage, you're focused on the downtown cores- density, density, density.

00:08:40.309 --> 00:08:44.809
And those are exactly the areas that were most impacted by these office shutdowns.

00:08:45.200 --> 00:08:55.860
So, now you're in 2020 or so, and the core assumption of your business around Dan's offices totally changes, you have to find product-market fit again.

00:08:55.860 --> 00:09:01.860
Maybe just walk us through what happens in this time period, and where you ended up shifting towards.

00:09:02.460 --> 00:09:06.360
Yes, great question, Pablo.

00:09:06.931 --> 00:09:09.870
COVID did flip the world around for us.

00:09:09.870 --> 00:09:27.480
Our greatest strength, part of this inset around coverage was a secondary insight that office buildings were a special entry point into look because the workday routine was basically what Ritual...

00:09:27.480 --> 00:09:37.830
it tends to be small order sizes, they're not great for delivery, most people, have food courts to buy or merchants at the bottom of their buildings that they want to transact from.

00:09:37.831 --> 00:09:40.740
And it's really about removing the friction out of the workday.

00:09:40.740 --> 00:09:45.389
Our growth with office workers was viral in nature.

00:09:45.390 --> 00:09:54.299
We had a very dense set of merchants in a very small area, so we could support them really well, as opposed to them being spread out across large areas.

00:09:54.360 --> 00:09:59.759
We really designed the entire company, the playbook, our go-to-market, everything was designed around capture.

00:09:59.760 --> 00:10:04.769
Capture the office area, and then expand outwards to a city from there.

00:10:05.490 --> 00:10:11.190
It's almost like that was like the cream of the city that we were trying to get focused on.

00:10:11.191 --> 00:10:18.779
And in 2019, we took on a very large expansion, and launched office areas across almost seven countries, 60 cities.

00:10:18.809 --> 00:10:24.779
It was an enormous expansion because of how well things had gone in previous years.

00:10:24.809 --> 00:10:29.009
And then March 2020, the global shutdown of offices happened, they're still closed.

00:10:29.010 --> 00:10:35.190
And that was tough on the entire company and the space, even more so.

00:10:35.191 --> 00:10:48.360
And so, we've kind of had to go through a version of what we had to do in 2014, and we're a much larger company, and we thought,"there's still a meaningful business that'd be supported" but a few things happened.

00:10:48.360 --> 00:10:53.370
One was, we went from a company where 80 to 90% of our revenue came from offices.

00:10:53.700 --> 00:11:02.250
And now, even though we took a step back in revenue after offices shut down, and we've been rebuilding.

00:11:02.250 --> 00:11:08.549
Now 90% of our revenue comes from residential because offices are closed.

00:11:09.721 --> 00:11:18.230
So, people, their patterns, their workday patterns of buying have shifted where they are still making those purchases, but they're making it near where they live, not where they work.

00:11:18.230 --> 00:11:22.200
And so, we've had to do a lot of work to lean into that.

00:11:22.110 --> 00:11:30.120
And it's a more difficult problem because the one nice thing when you're a small team, and you're trying to get to product-market fit, is that things cannot be time pressured.

00:11:30.571 --> 00:11:43.799
The challenge when you're in market with a product, and you've got a fully functioning business that's scaling and the rug gets pulled out from under you, you have a lot of time pressure to figure it out, you can't take...

00:11:43.830 --> 00:11:47.639
So, I think it's tougher.

00:11:47.640 --> 00:11:50.370
You have to make bets quickly.

00:11:50.370 --> 00:12:02.559
It's kind of like wartime, it's more important to make a decision and move it forward than it is to perfectly get the right decision, get the decision right every time, and you've got to move very fast.

00:12:02.051 --> 00:12:04.870
And that's what we've done over the last 18 months.

00:12:04.870 --> 00:12:07.090
And it's basically been, we've done it in two ways.

00:12:07.091 --> 00:12:12.490
One is we've leaned into the fact that growth has now shifted from offices into residential.

00:12:12.490 --> 00:12:22.570
And we've built a series of products to expand our presence from just offices, including, building online ordering capability for restaurants that they can deploy on their websites.

00:12:22.571 --> 00:12:31.059
It's still very aligned with our original mission of enabling local businesses to go digital, but a different way.

00:12:31.690 --> 00:12:36.370
And frankly a much broader way now.

00:12:36.701 --> 00:12:47.710
But also we see it's starting to happen in Q4, and we think that it's going to happen over the next year, but it's going to be different in the sense that people have now been trained on digital ordering.

00:12:47.711 --> 00:12:58.240
And so, what used to be something that we were almost the early company showing people a new way of doing things and now becomes the default.

00:12:58.059 --> 00:13:07.389
A lot of buildings now see this as a necessary amenity, not an optional thing like it was before.

00:13:08.740 --> 00:13:17.860
And so, I think the good in it for us is that, what uses to be an optional thing is becoming almost a mandatory amenity in buildings.

00:13:18.341 --> 00:13:26.139
Also had to change our products to more deeply embed ourselves into thousands of office buildings, et cetera, et cetera.

00:13:26.140 --> 00:13:35.409
We're doing a lot of work to...It's hard to predict exactly what the world's going to look like when we're on the other side of this, but the data is basically the work we're doing right now.

00:13:35.799 --> 00:13:36.220
Awesome.

00:13:36.700 --> 00:13:40.509
All right, we'll cap it off there, give a quick summary.

00:13:40.510 --> 00:13:46.779
We went through how, while working at Google, you started to see some interesting insights around local and digitization.

00:13:47.230 --> 00:13:55.629
And you ended up deciding to build Ritual a pickup app, how you devise the first entry point experiment.

00:13:55.630 --> 00:14:06.730
And as you said, that luckily that one went very, very well, and ended up getting you to scale across 60 cities worldwide, in many different continents.

00:14:06.730 --> 00:14:24.159
Only to then have COVID come and really totally changed the game, and force you to go back to first principles, reorient your company, and honestly, come back stronger than ever as COVID fades away and as you now have more presence, even in the suburbs.

00:14:24.160 --> 00:14:33.399
So, quite an incredible journey, I'm sure founders will learn a lot listening to the way you describe your thinking and all the steps you took.

00:14:33.400 --> 00:14:34.240
So, thanks a lot, Ray.

00:14:34.259 --> 00:14:35.379
Really appreciate having you on the show.

00:14:36.700 --> 00:14:36.700
Thanks, Pablo.

00:14:36.700 --> 00:14:36.700
It was a fun conversation.

00:14:39.190 --> 00:14:40.210
Thanks so much for listening.

00:14:40.509 --> 00:14:43.179
If you want to see more content, check out pmf.show.