Growing a SaaS App for the Film Industry With Rigorous Experimentation

Hello! What's your background, and what are you working on?

Hi! We are Alan Grow and John "Pliny" Eremic. We built Endcrawl.com.

Endcrawl is a SaaS app for making end credits. You know — all those names and logos that scroll by at the end of the movie.

More than 1,000 film productions have used Endcrawl so far. Some of them you've heard of, like Moonlight, which famously bested La La Land for last year's Best Picture Oscar; or nominees this year like The Big Sick, Mudbound, and Lady Bird.

One-third of the entire Sundance 2018 lineup used Endcrawl, too.

But we are just as excited about the 1,000 other movies driving Endcrawl that you've probably never heard of. Our ability to sell deep into the indie film market is a big part of our story.

What motivated you to get started with Endcrawl.com?

Pliny ran a post production company for eight years, where he worked on scores of movies, including "The Great Gatsby."

During that time he pegged many areas where film productions were — in Alan's phrasing — "poorly automated chaos engines."

But while many of those were too small to build a business around, end credits jumped out as a hair-on-fire problem. Every movie needed them. The going rate was $1,500 to $10,000 a pop. And both clients and vendors dreaded the workflow: long email chains, Word and Excel attachments, unstructured data, and a slow and buggy feedback loop.

The problem space lent itself uniquely to a SaaS. For a hot minute we considered a plugin for NLEs (video editing software). But we felt that desktop software would limit our customer base. It would also fail to address core problems like collaboration and turnaround time.

We knew we would be delivering massive amounts of video to customers. For example, a 5-minute end credits sequence in 4k resolution weighs in at about 250GB, unzipped. So our back end would have to be rock solid. Alan brought a strong full-stack and DevOps background to the table. He'd done his time with the tech giants (Microsoft, Bloomberg) and had recently co-founded art-tech company iLuminate.

One night at an industry holiday party, we had one too many drinks and finally talked ourselves into building the damned thing. Cocktail table napkin sketches may have been involved.

What went into building the initial product?

Our MVP was a Perl script that Pliny built at his breakfast table on a Saturday morning. (Groan. —Alan, Technical Co-founder)

In fact, we might be a textbook Lean Startup case study. We're not dogmatic about Lean, but it has served us extraordinarily well.

Concierge Stage

The Perl script became our concierge MVP. It worked like this:

• We gave customers a Google Sheet with their end credits formatted to our spec.
• They could edit the sheet to their hearts' content.
• When they wanted a new "render" (a video output), they would email us.
• We manually exported their Google Sheet to CSV.
• Then we manually ran that CSV through the Perl script.
• We manually uploaded the resulting files to Dropbox.
• We manually emailed the customer their download link.

That process took about 15 minutes. But filmmakers were used to waiting between 6 and 24 hours!

The core UX we were going for here was:

1. Customer controls their data.
2. Video outputs are fast, fast, fast.
3. Unlimited do-overs.

In those days we often felt like Food On The Table founders featured in the Lean Startup book, personally preparing meal plans for individual customers. This concierge treatment wasn't going to scale, but it helped us learn and validate some of our core assumptions.

And this resonated. In short order, we had acquired our first two dozen customers. We knew we had something by the tail. But just how big it was, we weren't sure.

Email-first Stage

The next step was to automate some of the process and try to scale up. So we moved ahead to the "Email-First" playbook.

In this phase, customers could request renders over email. Instead of founders scrambling, we had several Unix programs and a makefile handling each request.

This went over well, too.

In fact, we completed just shy of 100 productions — who were happy to pay $500 to interact with an API over email — before we even rolled out the first iteration of our UI.

How have you attracted users and grown Endcrawl.com?

Once the Perl MVP existed, we cobbled together a landing page with a WuFoo form and sent an email blast to Pliny's filmmaker contacts. That was the extent of our launch.

Early growth was 100% organic. But what we found is that as we progressed, growth got harder — not easier. There is this founder fallacy that as you grow, everything just starts flowing downhill. We found the opposite to be true: once all of the low-hanging fruit has been picked, the real work of selling starts.

If you're at this stage with your company, take heart. It doesn't (necessarily) mean the market isn't there. It just means you actually have to get good at marketing.

The best thing we've ever heard on this topic is Gail Goodman's presentation "The Long Slow SaaS Ramp of Death." It's required viewing.

We could probably write a book on all of the marketing experiments we've run since. Here are some highlights.

Worked: Native Ads

We place our name and logo in all new Endcrawl projects. Customers can remove them if they like of course, but the power of default is strong. As a result, our name and logo are sprinkled across hundreds of films and will be seen for many years to come.

We encourage all founders to think creatively about what constitutes a "native ad" in their space.

Worked: Customer Interviews and Demos

One-on-one meetings and demos are still key for us. They don't always lead to a direct sale, but they increase our word-of-mouth. Peer referrals are a huge driver in our industry. So we've learned to focus efforts on things that increase that diffusion coefficient.

We also conduct interviews with new, recently converted customers. Meeting so many filmmakers is a great networking opportunity in its own right, and the personal connection makes our customers more likely to recommend us. We always learn something new when we ask questions like "What didn't work for you at all?", "How did you hear about us?" and "Where else might we have reached you?"

Worked: Professional Groups and Meetups

It helps to be a good citizen in your niche. We have worked with, and sponsored, local industry groups like the Blue Collar Post Collective; and Pliny does guest lecture spots at places like Columbia University, Feirstein, and the Manhattan Edit Workshop. These are all brand marketing exercises as well.

Worked: Content Marketing

Our space is saturated with how-to filmmaking linkbait. What worked for us was highly differentiated long-form content written in a clear and unique voice. For example, Pliny here doing his best Ben Thompson impression. Occasionally, one of these articles even pops on Hacker News.

Worked: Swag

Swag can be a vanity exercise, but it can also be incredibly effective. As in tech, logo tees are common work attire in our business. We know exactly where our target customers (producers) stand when they're on a movie set, and we distribute our apparel to crews in order to maximize visibility.

We also keep our biggest advocates well-supplied. A little love there goes a long way.

One more note on apparel: be inclusive! It's not hard. We offer three different styles. If you're considering printing tees, we strongly recommend reading this first.

Worked: Paying attention to our Google Analytics

We take note of traffic spikes and do our best to chase them down. Once it was a private Facebook group that sent a ton of traffic our way. Another time, it was a small but highly-targeted industry newsletter. We sponsor that newsletter now.

Worked: Twitter!

Hard to believe, but Twitter once had a stronger lookalike product than Facebook. We got some good traction on that platform early on, like with this promoted Tweet.

We also ran some fun experiments leveraging Mechanical Turk. We would use it to generate Twitter lists of filmmakers who had just been added to a new festival lineup, then engage with those accounts. It worked!

But diminishing returns set in. Behavior and preferences change over time on social media platforms. And, even more simply, there just aren't enough people on Twitter.

Didn't work: Facebook

We've gotten pretty good at building custom audiences on Facebook. Unfortunately, the results have been weak. But our customers simply don't like being advertised to on Facebook. Even more to the point: they wouldn't dream of picking a professional tool based on a Facebook ad.

How did we learn this? In all of those one-on-one interviews.

Didn't work: Display Ads

Our industry has a few online communities, but most of those aggregators have not thought past banner ads. It's a damned shame. We still give it a shot from time to time when we see a new angle. But the display ads perform about as well as you'd expect in 2018.

Didn't work: Direct Response

A broader pattern we noticed was that Direct Response advertising had very mixed results.

Endcrawl isn't a service that you need 365 days a year. Customers might hear about us, but generally sign up the moment they need us and not a second sooner.

Realizing this forced us to put more resources into brand advertising. Most of the founder advice you read assumes Direct Response models and subscription pricing. Be cautious if your company doesn't fit that template.

Everybody — us included — wishes for an easy, growth-hacky flywheel. Wouldn't it be nice if we could just plug dollars into Facebook and sit back while the signups roll in?

Don't count on this. In our experience marketing and sales are not set-it-and-forget-it. Selling is a core, existential founder skill.

What's your business model, and how have you grown your revenue?

We've found that pricing is inextricable from UX.

Our core value prop revolves around unlimited do-overs, and owning the process. Therefore:

• If we capped video outputs, we would disincentivize usage. We want customers to make as many renders as they like, because that is such a huge differentiator for us.
• If we charged subscription fees, we would incentivize our customers to sign up late and try to crank out their credits in a few days. At a minimum, this would place undue stress on our support team. And customers would prefer to tinker with their end credits for as long as possible.
• If we charged per user, it would incentivize account sharing. End credits have many stakeholders — it's better when everyone is on board. And more users means greater virality for us, too.

So we decided to charge a simple, flat fee per project.

Each project has unlimited users, unlimited usage, and unlimited renders. Low-res preview videos are free, and projects upgrade for either $499 (HD and 2k) or $999 (UHD and 4k).

We started charging on day one. There is no value in learning that someone would gladly use your product for free.

Now, our model may be an anomaly. But the takeaway is: make sure your pricing doesn't work at cross purposes with your core user experience. Pay attention to what incentives your pricing model creates, and keep those incentives aligned with your value prop.

What are your goals for the future?

The goal is and always was freedom. First, freedom to conduct business according to our values. More broadly, to live life on our terms. That includes financial independence of course, but it's also about the freedom to structure our lives and routines in healthier ways. The 9-to-5 thing runs roughshod over a lot of natural cadences.

Some startups end up being a golden calf, always demanding more tribute from their founders. We are trying to build a company that exists to serve us, our employees, and our community — not the other way around.

What we've built so far is not just a SaaS, but a sales pipeline, a brand, and a strong founder relationship.

That plus a growing revenue stream give us a ton of optionality in life.

Or, to quote Jony Ive:

"I always think that there are two products at the end of a programme; there is the physical product or the service, the thing that you have managed to make, and then there is all that you have learned. The power of what you have learned enables you to do the next thing and it enables you to do the next thing better."

—Jony Ive (source)

What are the biggest challenges you've faced and obstacles you've overcome? If you had to start over, what would you do differently?

Challenge: Part-time, different time zones

Alan and Pliny live in different time zones and have full-time day jobs. Running a distributed company part-time presents a unique set of challenges. Some ways we've addressed this:

• Face to face bonding and hackathon times.
• Thinking very hard and deliberately about our communications stack.
• Establishing a cadence: stand-ups, reasoning sessions, and "office hours" (times when we are both working at the same time).
• Setting clear expectations around response times for things like feedback or code review.
• Working out our written culture. A distributed culture needs a strong written culture. That means strong internal documentation, establishing style guides, agreeing about how we message to customers, internal shorthand and inside jokes, and knowing how to run a good Pull Request.

Challenge: Customer support

Early on we were able to handle support during nights and weekends. But we grew out of that.

The first challenge was just admitting to ourselves that we had to hire. It's perhaps a bit unique to be part-time founders running a support team that's open for business 12 hours per day.

That brought a new set of challenges: onboarding support reps, setting up a helpdesk system.

But frankly, this is one of the best things that happened to our company. Indie hacking can be a lonely business. You can get too far inside your own head. Our new team members made huge contributions to our culture. The Slack is way more fun now, too.

Tech "Regrets"

Looking back, we should have moved to AWS sooner. We used Dropbox for storage initially, but the idea of using shared Dropbox folders in the product never panned out. Dropbox isn't always reliable, and while it never bit us hard we should have thought through cloud storage more carefully.

Another big thing we procrastinated on was unsnowflaking our servers, and just generally getting our DevOps house in order. Again, while it never bit us, the danger of an extended outage posed an undue risk to our brand early on. The upshot: we wound up open-sourcing our zero-downtime deployer.

Every technical choice you make might feel like a mistake in the distant future. Accept it but be a minimalist. It's the best way to control the damage.

Mistake: Not selling harder, sooner

We should have been more aggressive about selling from day one. Growth is compound interest: the more you put in early, the more you will have later on.

One of the reasons we didn't sell harder, sooner: we were still getting comfortable with the whole "Lean" thing. Film industry peers told us "perception is everything" and that we'd be crazy to launch without a hyper-slick, fully-featured, and beautifully considered web app. We allowed ourselves to feel self-conscious about things like our Google Sheets integration, and being concierge/email-first.

In retrospect, we shouldn't have worried. Launch sooner, sell harder.

Have you found anything particularly helpful or advantageous?

Read "Traction"

We came across the book Traction, written by Duck Duck Go founder Gabriel Weinberg. That book is a swift kick in the pants: we came out of it with enough marketing experiment ideas to last for the next year or two. Highly recommended.

Perhaps the best takeaway: what drives growth in one stage of your company is not necessarily what will drive growth in the next. We probably spent too much time trying to wring blood from the Twitter stone, for example, when we should have moved on.

Keep an eye on your traction channels. Enjoy them while they last, abandon them when they run dry, and always keep sniffing out new ones.

Get a Board of Advisors

We only did this more recently. And we're kicking ourselves for not doing it sooner. BoA's aren't just for VC-funded startups.

Our BoA members provide sanity checks, strategic perspective, and key introductions. They are philosophically aligned, and they like us personally. In other words, they give a shit whether we sink or swim.

In funded startups, BoA members typically receive a small piece of equity. If you are too small-scale for that to make sense, there are plenty of other ways to structure the arrangement.

Everybody needs advice and mentoring. Get out there and ask someone wiser and more experienced for help. You'll be surprised how many people are absolutely delighted to say yes.

Check out Indie.vc

Indie.vc has done us all a service by providing a counter-narrative to the Sand Hill playbook. They have been a source of inspiration.

Full disclosure: we applied to Indie.vc and didn't make the final cut. But in hindsight, we think they made the right call. (No, really.) We're better off still 100% independent today.

In fact, to paraphrase a famous movie quote, you might say that the greatest trick that Bryce ever pulled was convincing us that we don't want his money.

What's your advice for indie hackers who are just starting out?

JFDI. The optionality alone is worth it.

We had no real sense of what the knock-on effects we would be when we started Endcrawl. But as soon as we put ourselves out there, things started happening. We met other founders. We grew new networks. We developed new skills. Doors opened.

To give a more concrete example: we didn't know our exact TAM (total available market) when we started. Turns out, nobody does. The global film market is distributed, amorphous, and very hard to map.

Endcrawl is a powerful market research tool in its own right. We have monetized student films, web series, nature docs from Australia, zombie kickstarters in Nebraska. It's been like sending a series of sonar pings into every corner of our market.

It might be an exaggeration to say that we have a better picture of our market than just about anybody else. Might.

You learn lots of new things when you start shipping.

Run the hell away from prestige customers

Some of the most common advice we've heard from film industry peers is "go after the big fish" and "pitch to studios." Terrible advice. Luckily, we didn't take it.

The idea makes sense for brick-and-mortar companies: if your resources are fixed, you need to chase higher hourly rates.

But for a software company, that logic is flipped: scaling resources is fairly trivial, so you need to concentrate on where you can scale. Those big fish — i.e. enterprise customers — are high-touch and don't scale well. (And just wait until you find out what it's going to cost in legal fees to paper that SLA.)

We've watched other startups get lost in the "enterprise sales" forest. It's a trap! That famous company logo you're dying to put on your landing page? That's pride messing with you. Ignore the siren song of prestige, and head down-market. Trust us, it's much nicer down there.

Work on that founder relationship.

Your co-founder relationship is like any other relationship: you have to work on it. Your company will live and die by its health.

We have regular stand-ups and "office hours", but we also hold separate check-ins that we call reasoning sessions. Those are reserved for discussing high-level strategy, politics, personal and professional struggles, life goals. It's crucial for us to keep sharpening that saw and relating what we do on a tactical level to the kind of lives we wish to lead, the kind of people we want to be, and the kind of world we want to leave behind.

We've joked that our founder relationship is really a philosophical alliance, and the companies and products we build are just by-products of that alliance. But that really isn't too far from the truth.

Where can we go to learn more?

Check us out at: endcrawl.com.

On Twitter:

• @endcrawl
• @iampliny
• @alangrow

We'd love to hear from the community! We're here to give, and we're here to learn. Ask us anything.