High CPA? It could be a negative indicator of product market fit
At Wine List, Facebook CPAs confirmed one of our core suspicions. Here's how you can test if this is true for you too.
When I was the founder of Wine List, there was one thing I found more difficult than almost anything else.
The balance between needing to break through walls, and being able to accept when something wasn’t working.
Our Facebook CPAs were very high at various points. Was this a problem of the platform or something bigger? Was this in fact a sign that we didn’t have product market fit.
Today I'm looking at a few of the things we did, and how if you're at this existential fork in the road, you can borrow our technique for testing this properly.
Background
We closed our pre-seed in February 2020. We were an education-focused wine learning platform. Watch some online content, read some learning materials, do some interactive wine tasting, and drink great bottles at home.
This was accidentally the perfect covid product. And this reflected in acquisition during that first wave of covid. Facebook was netting us CPAs in the £30-40 range.
We won partnerships with the likes of Soho House, plus significant press coverage in The Evening Standard and Monocle.
Our blended CPA in April 2020 was £24.
This was our first proper month trying to acquire customers and so we had no baseline for what was normal. We also had no data on retention or lifetime value. And because of how early we were, our margins were basically 0%. Sidenote: for anyone wanting to get into wine: it's a basically margin-less game.
As covid ended, paid acquisition CPAs shot up, and retention edged down. A year later and the same product was at £100+ on Facebook.
My first problem: trying to other channels to make blended acquisition work
Partnerships are brilliant. So too is good, relevant PR. Both can drive short term bursts of acquisition to your product. They also have great multiplier effects on your overall acquisition if they gain search rank. (I.e. a good press review will be part of someone's searching on your brand).
They are both brilliant in broader strategic ways too. Good press leads to better negotiating in your operations and buying. It helps with hiring and profile building. And it really helps with fundraising. Ditto, great partnerships can help create even better ones. They help establish your brand values. And they're useful in other strategic decisions and growth plays.
But they are not a repeatable growth engine. By that, I mean they are not something where you can reliably increase output by increasing the input. If you go from spending £1k per month on PR to £50k per month on PR, you're unlikely to get 50x the benefit.
The biggest priority has to be getting your paid acquisition working
You don't choose your marketing channels - your product, market, and business model do. If you're selling a consumer product for £10-500, then 95% of the time, your growth engine will be paid acquisition or 'performance marketing.'
The vast majority of businesses get big thanks to just one channel. And so the most important thing you can do is to focus your energy in getting that channel spot on.
But why’s that? These other channels are a great way to blend down your CPA. But because of their lack of scalability it means you’re at the whims of your primary channel CPA. And if anyone’s ever thought ‘I wonder what would happen if we doubled our budget’ you’ll know how bad things can get.
Consider this scenario:
In this example, referral becomes slightly less efficient as we scale 43% down to 37%, and PR grows 50% – which is pretty high. But even so, with the core channel CPA blowing up, you’re blended CPA has gone up with it.
This is what happened at Wine List: core channel CPA blew up. And suddenly to keep blend in the same place, we needed to do so much more work with the other ‘free’ channels just to stay even.
Why high Facebook CPAs were an indicator of no product-market fit
We had to get Facebook working. The search volume for wine education was <100/month. There were no network effects. Our product was too cheap for sales or events channels to be reliable as scaling tools.
We, like most other consumer businesses, needed to get paid social working. TikTok hadn’t launched ads yet. Snap was for kids. Meta was our channel.
The idea of ‘product-channel fit’ is often seen as a separate piece of the puzzle of product-market fit. But instead I see it as a required checkbox in your search for PMF.
Our Facebook CPA got high. 4-5x higher post-covid as it was before it. What this revealed was there was a much smaller market of people who wanted the product than we’d hoped.
We spent the next 6-9 months running the Superhuman PMF process, customer interviewing our members, making adjustments to the product, and making operational changes to see if there was a world where we could support a £100 CPA.
But there was one thing we did in Q1 2021 which we could have tested far earlier and would have revealed an answer.
High CPAs can be okay
One thing to note is that a CPA is only too high if it is unsustainable for your LTV. I’ve seen £500 CPAs where there’s the cash to pay back, and the strong enough long term LTV to make it profitable. But at Wine List, we were aiming for CPAs many multiples lower than they were on Facebook.
How you can use Facebook to test if there’s the market there for your product or not
We ended up testing a semi-large pivot. Our proposition was learning about wine and sending you two bottles per month to taste. But to properly learn, you ideally need a wider variety of wines in smaller volumes. Therefore, we had the idea to test smaller 100ml samples of wines.
When we eventually launched the product, we put new ads on Meta.
While our core 2-bottle variant was running at £100 CPA on Meta, and the six bottle case variant at closer to £200, the new 100ml sampler version went in at £10 CPA.
This was one of the few ‘Eureka’ moments we had. The difference between the existing paid acquisition experience and this one, was night and day. Suddenly there was a market demand for the new thing we were trying to get rid of.
We spent five months from prototyping to shipped product. Five months of a two year lifespan as a business. What we should have done is borrow from the Lean Startup and spend a week getting mockups done, start selling and validate earlier.
High CPAs and a question mark over CPA? Take this January challenge
Given that time again, I’d have got the product designed physically – or maybe better yet, just fully CG-Ied – built the landing page and ads, and just started selling. We could have seen in probably two weeks what it took us five months to learn.
Meta Ads are a reflection of the market. If you’ve got good creative, clear JTBD proposition, and a good LP experience, you’ll be able to see at a high-level what the market wants. It won’t be your true CPA. But in the examples above I bet we could have got a £20-30 CPA with mockups alone.
Startups today don’t have the luxury of the 2010s VC market. We don’t get the opportunity to spend time building before we take off. The plane is in the air. It’s time to test like crazy.
So if you’ve got hypotheses on what other product lines look like or what a pivot could mean, then do the mockups, and do the test. Opportunity cost is too big a risk.
Closing thoughts
As you go through small iterative changes and CPA goes from £65 to £63 to £60, it’s easy to think you’re working on the big things – but if your business model is an order of magnitude away from making that CPA work, you need much bigger swings. It’s not being a good entrepreneur to keep running at a wall that you’ll never break through.
Ballpoint NIBs
This is the section where we talk about other interesting things you might want to check out.
Got brand strategy or comms needs? Friend of the agency Leslie Campisi has just launched a new agency, the LA-based Beginners alongside Christopher Hastings Farrell. Their target market? Those who start things. I used to work with Leslie in my first ever job, and she’s been helpful with comms advice to me ever since.
Follow us on Instagram. We’ve just kicked off a sprint pushing out more content on Instagram. We’re leaning heavier on the entertainment side (we hope) than on LinkedIn but still plenty to learn from. Give us a follow: @weareballpoint.
Google begins the end of cookies in Chrome. This time next year, Chrome won’t track any cookies at all and that journey has begun. It’s another nail in the coffin of the 2010s approach to digital marketing where every element was trackable. This will have an impact, but as with everything change is inevitable – and in that change is opportunity. Make sure to lean in.