logo New Idea

February 2022 Retro

I spent 4 months building up the confidence to talk to customers. Something I could have done on day-one. Here's what they told me.

In last month’s retro , I learned the hard way that my entrepreneurial intuition is far from calibrated. I spent time on things that felt necessary, but was actually just a distraction from talking to customers. Since then, I focussed on doing customer conversations (as defined by Rob Fitzpatrick in The Mom Test ).

The biggest takeaway being that, I didn’t need to build a damn thing to get insights from my target customers. The only advantage that building out AdSlicer gave me is the confidence that I knew how Facebook Ads worked. The 4 months I’ve spent actually building AdSlicer has just been a glorified self-esteem building exercise. Fuck.

Last month’s goals

  1. Talk to at least 1 Facebook Ads marketer (that I don’t already know) about AdSlicer.

    Final grade: 💯x3

Main events

Speaking to customers

I managed to chat to 3 Facebook Ads specialists, each of varying experience.

Initially, I wasn’t too sure how to get into conversation with these people. I tried writing a sincere post about a lonely software developer building out a Facebook Ads analysis tool. But a post like that was classified as self-promotion and it was quickly removed by the admins.

Then, Michael Lynch reminded me of a classic Jason Cohen customer acquisition tactic :

Jason Cohen's customer acquisition tactic An example of Jason Cohen’s cold email that he sent to 40 consultants.

The tactic in a nutshell:

  1. Tell them what your intentions are (i.e. to ultimately help them).
  2. Prove that your intentions are pure by offering to pay them for their time.
  3. Pray that they want the problem solved more than they want to be paid for the time it takes to talk about the problem.

So instead of making a self-promoting post, I made a post about looking for paid consultants to give me feedback on my Facebook Ads analysis tool.

That post got a lot of interest. Mostly low quality marketing spam, people who didn’t read past “paid consultant” and just copy-pasted their generic pitch.

I picked up the conversation with 8 people and ended up on a call with 2 of them.

My delivery of the tactic was obviously poor because in both cases the person wouldn’t meet with me without an upfront payment 🤦‍♂️ (Profit: -$100)

But fair enough a deal was a deal, I was just happy to get someone on a call.

While going through all the DM’s and FB friend requests I received from my posts, I remembered a friend request I got from someone off the back of my previously removed post. I started up a conversation and hit the jackpot:

A customer agreeing to talk to me for free The response I got when I finally asked an ideal customer if I could pay them to talk to me about their problems.

After one no-show, we did have our 30min meeting that turned into an hour long meeting. This call was way more valuable than the paid calls. The conversation was filled with phrases like “I hate X about FB ads” and “I wish I had Y for FB ads”. None of which were “I wish I could visualize my FB ad results” 🥲

But that’s the point of talking to customers right? To narrow the gap between assumptions & reality.

Me: You know when people keep saying “talk to your customers”?

Also me: Yes, of course, startup 101

Me: Dude! They literally mean talk to your customers.

Also me: Yes yes, but what if 🤔 ...

Me: No, shut your mouth! Talk 👏 to 👏 your 👏 customers 👏

To be honest, after these initial calls, I feel pretty defeated. I built the wrong thing. And I’m still not sure what the right thing is. I feel like the problem is way bigger (& nuanced) than I thought. The tools that already exist are huge & complex and I could never achieve feature parity.

There is a small glimmer of hope in the fact that despite the existence of these other complex tools, the problem still isn’t solved.

The main problems I’ve heard around Facebook Ads are:

  1. Client reporting.

    This seems to be a huge time suck for solo marketers (8-10 hours per month). However, the problem seems to already be solved for marketers working in bigger agencies who have invested in existing automation tools.

  2. Previously well performing ad campaigns just stop working.

    The ad campaign stalls for some reason and stops delivering results. The industry wisdom seems to be to duplicate the ad campaign and rerun it. Seeing as FB is a huge machine learning engine, I think that the campaign is reaching some kind of local minima and getting stuck. So I can see that rerunning the campaign at a different starting point (in time) can yield different results.

  3. Uncertainty around scaling an ad campaign.

    It seems like simply throwing more money at an already working ad campaign doesn’t necessarily mean better results. You also have to think about scaling your audience size, avoid audience ad fatigue, and scale the fulfillment of your product too.

  4. What to do when the ad metrics are bad.

    There seems to be a lot of guesswork happening when optimizing ads because there are simply too many variables to consider. This lack of confidence in the next step can be quite stressful. Having a basic decision framework could be useful to keep the optimizations structured.

I started out this journey into Facebook Ads because I needed to analyze the Facebook ad results for my own eCommerce store ( back in September ).

That has progressed into wanting to help my girlfriend (a Facebook Ads specialist) make more informed optimization decisions for her eCommerce clients (and others like her).

In order to solve an ad issue, you need to know what the issue is.

For example, Facebook just tells you that your purchase rate is X%.

It’s up to you (the marketer) to know that:

  • X% might be low for your industry.
  • X% might be low because customers aren’t finding the product in their size.
  • X% might be low because customers didn’t expect to pay extra for shipping.
  • X% might be low because your website takes 10s to load.

And the same goes for every other reported metric.

AdSlicer just shows you X% on a graph instead of the usual pivot table. It doesn’t tell you what the root cause might be. It gives you a broad indication which metrics seem low, but it doesn’t tell you why it might be low.

For an eCommerce store, the happy purchase path is almost always the same:

  1. View ad
  2. Click ad
  3. View landing page
  4. Add product to cart
  5. Initiate checkout
  6. Purchase

In order to address the “What to do when the ad metrics are bad” problem, I think having a running list of possible causes for each low metric might make marketers feel more on top of things 🤔

I guess I’ll need to talk to customers again and see what they think.

Time

My total time allocation has dropped by 17% compared to the previous month (~132h). Mainly because of my trip to the Philippines. I’m currently still in the Philippines.

It’s been difficult to control my time as much as I did in South Africa. I feel very reliant on public transport now that I don’t have my own car. Everything takes longer than I expect it to and there’s nothing I can do about it besides throw more money at the problem, which I don’t feel comfortable doing right now.

I’m happy to be out & about and seeing how the rest of the world is living.

Time allocation (2022-02)

This month’s goals

  1. Retry Jason Cohen’s customer acquisition tactic without going bankrupt.

    I think I need to target more experienced marketers who have more Facebook battle scars.

  2. Get at least 2 verbal commitments to purchase AdSlicer once it’s ready.

    I think I was too timid in my initial round of customer interviews. This time I want to bring up pricing and push them to make some kind of commitment (even if it’s just verbal).

  3. Get more involved with Sheena’s day-to-day ad optimizations to help inform AdSlicer features.

    I’m fortunate enough to have access to live ad accounts of real eComm clients. I need to figure out how AdSlicer can give Sheena superpowers.

Thanks for making it this far. If you enjoyed this update, it would personally help me out a lot if you could retweet my post:

Alternatively, you can leave your email address below and I'll email you my next monthly update: