Snappr: Building API-First, Last
From Self-Service Photography Marketplace to Visual Content Workflow Software
Welcome to the 162 newly Not Boring people who have joined us since Monday! Join 104,321 smart, curious folks by subscribing here:
Hi friends 👋,
I’ve been writing about web3 a lot. There are a bunch of reasons for that, but one of the main ones is that I’m just a business model nerd, and new tools create opportunities for new business models. But I love digging into innovative and clever models wherever I find them, and today’s focus, Snappr, is a perfect example.
It started as an on-demand self-service marketplace for professional photography, and by listening to its clients, it’s built an enterprise SaaS and API offering uniquely enabled by the self-service business. It’s a model and product roadmap that you probably couldn’t have drawn up ahead of time, but that makes perfect sense looking backwards. And it’s a case study in how to build an API-first business, last, once all of the right pieces are in place.
Today’s post is a Sponsored Deep Dive on a company that I invested in through Not Boring Capital. As always, I only write Sponsored Deep Dives on companies I would invest in, and in most cases, I actually have. You can read my thought process on deep dives here. I’ll always disclose my conflicts. No conflict, no interest.
Let’s get to it.
Snappr: Building API-First, Last
Every big API-first business looks like it’s just software – just a few lines of code – until you look under the hood. There, you’ll find all sorts of complexity. Abstracting it away is the API’s raison d’être.
Most API-first giants, like Stripe, Twilio, and Plaid, start with the software and realize that getting the software right actually requires a lot of undercover schlep.
Snappr, on the other hand, started with the hardest part first. It built an on-demand, self-service, B2C-centric photography marketplace before realizing that all of that work put it in the perfect position to build the software infrastructure for the lifeblood of the internet: visual content.
Other companies have tried to build photography marketplaces or visual content workflow software to serve the enterprise, but starting upmarket has turned out to be a challenging approach. Counterintuitively, and surprisingly even to the Snappr team, the right approach seems to be to start by building out a self-service, on-demand photography marketplace serving mainly consumers, and then to work up to enterprises.
Somehow, in 2017, that marketplace opportunity was still wide open. In the era of on-demand everything, booking a photographer was still a nightmare. I experienced it myself.
If you’ve had to book a family photo shoot recently, this story might sound familiar. It was shocking to me. I almost … snapped.
In the fall of 2020, about a month after Dev was born, I decided it would be a nice idea to book a photographer to do a family photo shoot. I searched on Google, found someone with pretty good reviews nearby, and hopped on a call with her. She seemed nice and the price was OK, something like $300. A little expensive, but for a couple hours and hundreds of pictures of our little man that we’d get to use for Christmas cards and keep and cherish forever, not bad!
On the day of the shoot, we drove to a park all dressed up in our matching outfits. We met the photographer, we’ll call her Sally (not her real name), and she was lovely! Throughout the shoot, she showed us previews of the pictures, and they were great. What a day!
At the end of the session, as we were saying our goodbyes and getting ready to go home, Sally asked us when we wanted to schedule a Zoom to see the pictures. “Oh no, that’s OK!” I said, “It’s 2020, you should be able to just put them in a Dropbox and send them over to us, thanks!”
“I can’t do that,” said Sally, ominously, a shadow darkening her face.
Turns out, the price that we paid for the photo shoot… didn’t actually include the photos! That was separate. We’d have to buy the prints. “Well, what if we just want the digital versions?” I inquired. Nope. We’d only get the digital version of whichever physical prints we bought. And each physical print cost a minimum of $150!!!
I don’t get mad often, but I was fucking fuming. I went home and re-checked the price list and my emails with Sally. Neither the price list nor Sally ever said that we would have to pay per photo to get digital access to the photos. We’d been bamboozled!
It was a racket, but what could you do, not keep those memories? I gave in and bought four prints and a 25 pack of Christmas cards. It cost $799.69 (nice, but not nice), in addition to the $300 we paid upfront. I even kept the receipt to prove it. I never keep the receipt.
It was a jarring experience, partially because it’s such an uncommon one at this point. Today, practically everything that consumers spend money on is mediated through an online platform with full price transparency. We complain about Airbnb’s cleaning fees or DoorDash’s delivery fees, but we know about them before hitting the buy button. Price transparency is table stakes.
So last July, I was in the right state of mind when Basis Set’s Lan Xuezhao, a Not Boring Capital LP, asked me if I wanted to meet Matt Schiller, the CEO of one of the fastest-growing and most fascinating businesses in her portfolio: Snappr.
Snappr started its life in the US in 2017 as a B2C photography marketplace. I don’t know how I missed them when I was doing my photographer search, but it was exactly what I’d needed. I mean, check this out:
Online booking! Price transparency! All-inclusive!
If I’d booked with Snappr, the whole thing would have cost me less than a quarter of what Sally cost. As a consumer, that’s exactly what I needed.
But as a professional newsletter writer who needs to write about new models, a B2C photography marketplace probably wouldn’t be enough to do the trick.
The web2 marketplace model is tried and true. I even dunked on it a little bit when I wrote about Braintrust last week. If Snappr were following the familiar B2C Marketplace playbook – build up supply and demand, win the market, raise take rates by raising fees on both sides – we probably wouldn’t be here talking about it today.
But we are here, so what’s up?
Snappr is running a B2C2B2PLG model to perfection and writing a new playbook.
Haven’t heard of the B2C2B2PLG model before? That’s probably because I just made it up. I should explain.
Snappr has two business lines: Snappr Shoots and Snappr Workflows. Snapper Shoots is the photography marketplace business that’s open access, self-service, and asks customers to book and pay for one shoot at a time. It’s a consumer-like experience whose revenue is roughly 2/3 consumer and 1/3 businesses (mainly SMB). Snappr Workflows is the visual content SaaS and infrastructure offering for enterprises. With Workflows, enterprises can automate their visual content pipeline from beginning to end with Snappr’s APIs, and source, edit, and manage visual content in bulk.
Workflows is a bold expansion. In four years, Snappr has become the leading on-demand photography marketplace:
It’s processed over 120 million photos and edited more than 40 million of them.
Its network covers 90% of the US population, 95% of Canada, 88% of the UK, and 77% of Australia, among others
Its revenue has grown at a 14% compound monthly growth rate (CMGR) since it started earning money.
Its gross merchandise value (GMV) is in the tens of millions of dollars, and it captures a strong percentage of that as revenue.
The company has raised $13 million and has only spent a fraction of it to accomplish all that.
The majority of the company’s revenue now comes from enterprise (two-thirds of the company’s revenue now comes from B2B use-cases more broadly).
The company could have kept its head down and built a really nice photography marketplace business. Instead, it’s taking all of the resources it’s built up and launching Snappr Workflows, its SaaS and API product aimed at the enterprise clients who demand it and the mid-market clients Snappr hasn’t yet been able to serve.
The secret to Snappr’s success is that the Shoots business feeds the Workflows business, in four main ways:
Solves the Chicken and Egg Problem for Enterprises. Enterprises need national supply (photographers and editors) overnight, which is impossible to build on its own. When they call, you need to be ready, or you blow your chance. So you might try to build up the supply network ahead of demand, but then supply gets burned waiting around with no work. Self-service demand fills the gap.
Self-Service Increases Liquidity. The higher density the bookings are in an area, the stronger the network effects. There’s more supply density, less photographer travel time, and a greater chance of matching the right photographer to the right booking.
Keeps Photographers Happy. Most photographers like being able to mix it up and do B2C and enterprise shoots in the same day.
Shoots is the Biggest Channel for Workflows. 48% of Workflows enterprise customers today made a self-service booking with Shoots before becoming Workflows customers.
This last point is where that lovely acronym – B2C2B2PLG - comes from: people who book Shoots, both as consumers and for their businesses, end up becoming the fuel for Snappr’s Workflows product-led growth engine and often grow into larger enterprise accounts.
And Workflows is a game changer for Snappr’s business. It turns the product from on-demand to API-first, like a Zapier for the visual content pipeline. Snappr’s story is an accidental playbook for building an API-first business, backwards. So today, we’ll dive into the evolution by covering:
Started Down Under Now We Here
Was the Picture Clear from the Beginning?
Snapping into Focus
So grab a Fosters, throw some shrimp on the barbie, and let’s hop to a land down under.
(I’m very sorry.)
Started Down Under Now We Here
Somehow, certain pockets of the world become really good at producing things that, on the surface, have nothing to do with their environs.
Japan is unmatched at turning out valuable media franchises, with five of the top 10 highest grossing of all time, including the #1 and #2 spots.
The Nordics produce a disproportionate share of hit video games, including Candy Crush and Angry Birds. Goals is next. (They’re also excellent at EDM).
Australia’s superpower may be visual content software.
Sitting in sixth place of the world’s most valuable startups with a $40 billion valuation – right next to giants like Bytedance, SpaceX, Stripe, and Epic – is a graphic design platform founded in 2013 in Perth, Australia: Canva.
Snappr is gunning to be the next Aussie visual content software atop the global leaderboard.
Matt Schiller was born in the Outback on a little farm 15 minutes outside of a little town of 3,000 people called Hay. Growing up, Matt wanted to be a doctor. When it was time for uni, he went to The University of New South Wales (UNSW Sydney), which matter-of-factly calls itself “One of the best universities in Australia” in its Google preview.
At UNSW Syndey, Matt studied philosophy and medicine, graduating with a BA in Philosophy, a Bachelor of Medicine, and a Bachelor of Surgery. He also took an elective term in Neurosurgery at Oxford, and received a graduate certificate in tertiary teaching from Curtin University. Slacker.
With all that education, the moment during his educational journey that had the biggest impact on Matt’s career was the last one: seeing his mum’s emotional reaction to his med school graduation pictures. Hold that thought.
Before going to residency, Matt decided to do two non-doctory things:
He founded GownTown, now “Australia’s largest online university graduation retailer”
He took a “12 month break” to follow his wave of friends going into consulting.
For his “gap year,” Matt joined McKinsey to see the world, travel around, and work in healthcare from the business side. By year two, he deferred residency again and started doing non-medical consulting work. By year three, when McKinsey sent him to the US to work on tech and education clients, he realized that he was never going to go back to being a doctor.
Instead, he refocused on GownTown, and realized, both thinking like a consultant and remembering his mom’s graduation emotions, that there was an even bigger adjacent opportunity: graduation photography.
Snappr was born.
Was the Picture Clear from the Start?
Let’s pause here for a second.
Whenever I see a strategy that looks brilliantly constructed, I try to figure out whether it was all part of the plan from the beginning. Sometimes, companies go to great lengths to tell a coherent narrative that makes the company’s history look like the unfolding of a masterfully orchestrated master plan. That’s rarely the case.
In Snappr’s case, all of the ingredients were there for a smart consultant to piece together.
There was the market trend. Digital photos were becoming a bigger part of both consumers’ and businesses’ identities. The internet had become more visual. Airbnb was a case study for the value of high-quality photography. People and businesses alike would need, and be willing to pay for, higher-quality photos.
Then there was the market structure. I didn’t know this until Matt told me, but there was only one services industry more fragmented pre-marketplaces than the photography industry: taxis. Putting a digital marketplace on top of that industry resulted in multiple multi-deca-billion dollar companies, including Uber, Lyft, and Didi.
Unlike those companies, though, who would need to await the arrival of self-driving cars to produce software margins, the margin expansion opportunities for photography were manifold, from editing to storage to asset management to distribution.
You’d just need to grab the consumer market as a wedge, build up liquidity, write some code, sell the whole package to enterprises, and Bob’s your uncle.
Well, I got to ask Matt directly whether the plan all along was to go from B2C to B2B to SaaS and infrastructure.
The answer was a resounding, Australian-accented “no.” (noyr?)
Through his experience running GownTown, Matt just saw that there might be an opportunity to build a marketplace for graduation photographers. Then, McKinsey moved him to the US, and when he got here, he said that he started thinking bigger. Instead of just graduation photography, Snappr wanted to make it easy to book any professional photographer on-demand. They applied and got into Y Combinator.
At that point, what the B2C product would become was actually pretty clear. In the February 2017 TechCrunch article announcing Snappr’s YC acceptance, they laid out the customer value prop that has remained central to the Shoots product today: on-demand, affordable, high-quality, and transparent.
The article also mentioned that Snappr only accepts 5% of the photographers who apply to work on the platform. When I spoke to Matt recently, he said that they still only accept one out of ten photographers who apply in New York City after a four-stage vetting process that includes looking at a photographer’s portfolio and equipment, piece-by-piece. Photographers who do get approved get approved on a use-case-by-use-case basis: I might be really good at taking pictures of burgers but horrible at taking pictures of babies, so Snappr won’t let me near the little ones.
Matt told me that particularly in big cities, most of the photographers come inbound thanks to Snappr’s strong brand recognition and word of mouth from other photographers. They’re attracted by what they don’t have to do – sales and marketing, back and forth with clients, billing – and by the higher effective rates they make through Snappr.
Importantly for both sides, Snappr tips the traditional photography IP model on its head. Snappr sometimes charges more upfront than some other photographers, like our friend Sally, but that price includes everything. So instead of charging some low rate to lure you in and then charging for the photos, Snappr does what any logical person in the whole entire world would imagine they’d do: they give one price upfront (plus a small trust & safety fee). That price comes with all of the photos from the shoot, as it should. Look:
B2C was the plan coming out of YC in early 2017, and it was the plan when Snappr announced a $2M Seed round from firms including Airtree Ventures (Australia’s top venture fund) and angels like Zynga’s Justin Waldron and Google Maps’ Lars Rasmussen in September 2017.
At the time, TechCrunch explained why people were willing to pay for photography:
In today’s Instagram-first world, good photos are more important than ever. And like it or not, for some people a big part of spending money on fancy meals and experiences is the chance to share it with their social networks. So if you’re that person, why not spend an extra $59 on a Snappr photographer to make sure the photos are top notch?
(FYI: the lowest price package I can find now is $89… thanks inflation.)
Of the 5,000 or so bookings Snappr had done at the time, the most popular types were internal company events (SMB B2B) or family birthday parties (B2C), but “They’ve also had demand for profile picture photoshoots for LinkedIn or dating apps like Bumble and Tinder.”
(You laugh, but I’ve been using this same headshot that ROOM took for an article in 2018 ever since, and I would gladly pay someone on Snappr to take a new one (and will as soon as I lose just like 2-10 more pounds).)
Anyway, the first few quarters that Snappr generated revenue, it all came from the self-service business. The company started doing some real revenue in Q2 2018, doubled it by Q4 2018. All of it still from the self-service product.
And then something curious happened: enterprises started beating down their doors.
They pushed it off for a couple of quarters – the business wasn’t designed to do the kind of volume that enterprises needed to do – but this is a YC company after all, so finally, they took Paul Graham’s advice:
(FWIW the other half of the advice was “Don’t steal $3.6 billion worth of Bitcoin,” but not everyone listens.)
When Snappr did start talking to potential enterprise customers, they realized something incredible: the enterprise customers were often people who had used the product as consumers. They loved Snappr so much more than whatever their company was doing for photography that they wanted to use Snappr at work.
The team decided to run a test of a more ‘enterprisey’ product. They called it a “crappy MVP.” Customers loved it.
Enterprise usage started to become substantial in 2019. It was largely similar to Snappr’s consumer business, except that enterprises needed more photographers and centralized editing, so Snappr expanded its network and built an in-house photo editing operation.
Snappr’s enterprise business was a pleasant bonus. The unit economics and retention were incredible. Many enterprise customers came in, and still come in, through the self-service business, meaning that blended customer acquisition costs (CAC) are much lower than they would be without the self-service business. It was a nice cherry on top.
And then COVID hit, and the cherry became a lifesaver.
When COVID hit, as expected, self-service revenue fell off a cliff. People weren’t having parties, or in-person graduations, or doing much of anything that involved interacting with another person.
Matt saw the early signs and prepared Snappr to hunker down and weather the storm. The company laid off 25 employees. They prepared for the worst…
… and the worst never came. It was the opposite, in fact. While consumer bookings fell off a cliff, enterprise demand started to take off. It was puzzling at first, but it was obvious in hindsight: COVID, as we’ve all heard thousands of times, accelerated the shift to ecommerce. Product photography is the lifeblood of ecommerce. Everything that was offline moved online, and all of it needed pictures.
During COVID, enterprise became the majority of the business, and it’s remained that way since, even as the self-service segment has bounced back. Snappr re-hired all of the positions it eliminated, and added a bunch of new ones (there are tons of open roles today).
Turns out, the self-service business doesn’t just translate into enterprise business; it leads to investment, too. Lan at Basis Set told me how she came to lead Snappr’s A:
I found Snappr at my then 4-year-old friend Cora's party. I was surprised that a group of 4-year-olds had a professional photographer taking shots at a random gathering. It turns out that mom Theresa works at Airbnb and had found out about this new company called Snappr, which provides an Airbnb-like experience (Theresa also later invested in the company). The next day, I got a beautifully designed album and a very easy to use intelligent workflow. I was very impressed and reached out to Matt cold.
There’s a lesson in there: a great consumer product as part of a larger offering can be marketing that pays you.
Snappr emerged from 2020 with a pile of cash and sitting on a gold mine, with a new twist on a bunch of popular models.
Making a product so good that the individual employees inside a company want to use it enough that they push their companies to buy it isn’t new. It goes by many names: bottoms-up sales, product-led growth (PLG), B2C2B (business-to-consumer-to-business). It’s the model that propelled Slack, Dropbox, Notion, Figma, and Zoom to great heights, and as a result, is the model that many, many SaaS companies employ today.
Putting a structure around listening to product-led customers and turning them into actionable leads in order to grow accounts isn’t new either: it’s called product-led sales. Twilio famously didn’t think they needed a sales team because of their PLG motion until they lost the Uber account. They quickly learned to turn self-serve leads into sales targets.
Building something internally and then selling it to customers isn’t new: it’s how Amazon ended up with AWS and how Privacy.com ended up launching Lithic.
Consumers pushing their companies to officially use a marketplace product that they use and love as consumers isn’t new: Uber for Business and Airbnb for Business are real things.
Treating the individuals inside a company as ambassadors for a consumer product isn’t new: SeamlessWeb grew its B2C business as its clients’ employees used it, and shared it, personally (although Not Boring Capital portfolio company Sharebite is eating its lunch in the office now).
These last two shouldn’t be surprising: businesses are full of consumers, and most consumers work for businesses. Serve their needs in one context, and there may be an opportunity in the other.
But where Snappr has ended up by following its customers feels new, like a combination of all of the above:
Consumers used Snappr’s on-demand product.
They loved it enough to use it at work, too.
Snappr built new internal tools to handle the increased demand from companies.
Snappr sells those tools to enterprise customers as a combination SaaS and API-first product.
Enterprises start to use the software alone in order to work with existing or user-generated content.
Here’s how Snappr describes it visually:
Enterprise clients expanded Snappr’s market dramatically: from photography to visual content. Snappr estimates that the market size for visual marketing content is roughly $300 billion.
Enterprise clients also pulled new products out of Snappr. The company had already wedged itself into the beginning of a very fragmented visual content pipeline:
Humans book photo shoots, manually select images and videos from those shoots, edit them, drag them back and forth between files, and make sure they get to the right place in the app or on Instagram or wherever they’re needed.
Working with high-volume enterprise clients like Instacart, DoorDash, and Bring A Trailer having to do each one of those steps up to distribution on their behalf, Snappr built internal tools to make everything run more smoothly. They built:
Automated ways to trigger photographer bookings from business events
AI-powered photo editing tools to lighten the load on their in-house editing team
Software to manage quality control and distribution
Software to tie it all together and make it run smoothly.
Then last year, they started letting clients use pieces of that stack, and as of this month, they rolled it all out as a comprehensive product line: Snappr Workflows.
With the launch, Snappr is expanding from a labor-powered marketplace enabled by software into a software business backed by a labor-powered marketplace.
Now, the businesses that use Snappr for photography and editing on-demand can buy software from Snappr to handle their entire visual content workflow, from idea to live on site. Workflows integrate with the software clients already use, from Salesforce to Photoshop to Amazon S3. Clients can create workflows using their existing tools and Snappr’s. For example, for one of the big food delivery apps, the flow might look like:
new restaurant onboarded → book photo shoot → edit photos → post to Sheets for approval → post to app
Snappr can execute on all of it with a mixture of people and software. Every image comes back tagged and indexed.
Clients can also pick and choose which pieces they want to utilize. Some, for example, use Snappr to edit user generated content, mixing the authenticity of UGC with the quality of professionally edited photography. In that case, Snappr is mainly software – with a content pipeline and AI-powered editing augmented by human editors to get things just right.
Like Scale, over time, AI will be able to handle a larger percentage of the flow, improving Snappr’s margins. Already, more automation means that Snappr can charge lower prices, opening up the mid-market.
Five years into the journey, having survived COVID, the launch of Workflows represents a new chapter for Snappr, an opportunity to accelerate already eye-popping growth, and the last piece of a new business model.
Snapping into Focus
None of Snappr’s journey was planned to work out like this, but it’s working out nicely.
Snappr launched out of Y Combinator in 2017
It’s raised $13 million and spent a fraction of it
It evolved from a graduation photography company to a B2C marketplace for on-demand professional photography to a B2C and B2B marketplace for on-demand professional photography and editing to a SaaS and visual content infrastructure suite plus B2C and B2B marketplace for on-demand professional photography and editing.
Snappr believes that visual content is the electricity that powers the internet, and that it’s building the poles and wires.
Less than two years ago, Snappr was only a piece of that yellow box – sourcing – and even that was threatened by COVID. Today, it can orchestrate the whole flow: sourcing, manipulation, storage, asset management, enrichment, quality control, and distribution.
And it’s working really well.
Snappr is maintaining its 14% MoM growth rate, even as it grows off of a bigger base. As more of the bar turns red, it means that more of Snappr’s revenue is sweet, sweet SaaS revenue, and that it’s improving its margins without needing to extract more take rate from photographers and consumers.
Instead, Snappr has created a model that aligns incentives with its consumer customers and photographers.
Snappr Shoots is lead gen for Snappr Workflows. The effective LTV of a customer is so much higher than it would be in a traditional marketplace because each consumer comes with some probability of converting to a sticky, high-revenue, high-margin Workflows client. I would imagine that in some business-heavy markets, Snappr could actually subsidize the cost of Shoots well below market rates in order to acquire more Workflows customers. In that case, consumers win, Snappr wins, and Workflows clients, who can consolidate spend and receive a more cohesive product offering, win.
I always get to this point in the essay, my optimism firing on all cylinders, amazed by what these entrepreneurs have been able to accomplish, and forget to talk about the downsides or risks. In most cases, I don’t find the downsides that interesting. There’s that Tolstoy quote from Anna Karenina: “All happy families are alike; each unhappy family is unhappy in its own way.” I think it’s nearly the opposite for startups: “Nearly all startup risks are alike; each successful startup is successful in its own way.”
Companies face two main types of risk – market and execution – and if they’re lucky, a third: competition.
For Snappr, the risk is obvious: execution.
Building a marketplace business and operating it at scale is really hard. It’s hard to keep both sides of the market happy. It’s hard to maintain quality – what happens when there’s so much demand that Snappr needs to drop its quality bar from the top 10% of applicants to the top 20%?
Building good SaaS and APIs is really hard. Snappr is dealing with huge volumes of big files and thousands of unique client workflows and brand guidelines and quirks. Its DNA is as a marketplace company; success in one doesn’t guarantee success in the other.
Growing fast is really hard. There are product challenges and organizational challenges and everything gets complex really fast.
Building both businesses simultaneously while growing 14% every month, then, is an absurdly difficult challenge.
The flip side, though, is that if Snappr gets it all right, or gets enough of it right to keep moving and listening to customers and iterating, it has the opportunity to become the infrastructure for visual content.
Because even though Snappr was born as a marketplace, it has most of the hallmarks of a good API-first business. Photos are mission critical to modern digital businesses, but they’re non-core. They’re in the API-first Sweet Spot:
While API-first businesses look like pure software businesses on the surface, the big ones are just hiding a bunch of hard, dirty grunt work under the surface. As I wrote:
The magic of companies like Stripe and Twilio is that in addition to elegant software, they do the schlep work in the real world that other people don’t want to do. Stripe does software plus compliance, regulatory, risk, and bank partnerships. Twilio does software plus carrier and telco deals across the world, deliverability optimization, and unification of all customer communication touchpoints.
Snappr does a photography marketplace, photo editing, logistics, storage plus software.
They’ve come at it from the opposite direction – for similar reasons, I think, that toilets flush backwards in Australia – but the opportunity is every bit as massive.
What Twilio did for communications, Segment did for customer analytics, and Plaid did for financial data, Snappr wants to do for visual content.
Picture this: from a need for visual content to perfectly edited photos in just a few lines of code. Say cheese.
How did you like this week’s Not Boring? Your feedback helps me make this great.
Thanks for reading and see you on Monday,