Thread by @dunkhippo33

Today's tweet storm is on the strategy for testing customer acquisition channels.

Should you try a lot of channels? Should you try to make them all work? Which ones do you double down on? What happens if they stop working?

Read on >>

1) Yesterday I referenced a conversation w a portfolio co that is seeing great results w 1 cust acq channel.

Today I had a call w them about this strategy as well as what to do about their other cust acq channels.

https://twitter.com/dunkhippo33/status/1379646175016415232?s=20

2) At a high level, the most *ideal* situation is that you have just 1 customer acquisition method & channel. 1 playbook. People specialize & focus on the same thing day in & out.

That's the ideal. It doesn't work out that way, but that's what you hope will happen.

Why?

3) When you think about a machine, efficiency comes from doing things the same way over & over. You form a process. You execute that super well. You optimize it. You introduce as few variables to keep consistency.

The same applies to cust acq. You want to build a machine.

4) Put another way, here are things that get in the way of machine efficiency:

-New cust acq channels
-New playbooks introduced
-New team members
-New tasks for team members

You create inefficiencies & chaos w/ new things. More things can go wrong.

5) So that's why in an ideal world, you want to keep everything as simple, streamlined, and consistent as possible in customer acq.

The problem is in the beginning you don't know what is most optimal.

6) So in the beginning, startups try many cust acq channels.

In the case of our portfolio co, they are trying outbound sales. And various ad channels. And partnerships for distribution. Etc.

Many of these work decently well.

Of those, they have 1 ad channel that works best.

7) So what should they do? They should double down on that one ad channel!

They should be a bit more aggressive on CAC (let it go up a bit) since they can afford a longer payback period. They should monitor quality to make sure their new customer cohorts are still retained well.

8) They should keep doing this -- even take out a startup loan if needed to keep pouring money in -- until they hit a saturation pt. This is the pt where they cannot get anymore customers at that max CAC / max payback they are comfortable with.

9) So what to do about other channels?

This is where it's tricky. In an ideal world, they just focus on that 1 channel.

But what happens when that 1 channel saturates? You hit a wall. And that day will happen - it's just a matter of when.

10) When you hit that wall, you need another plan -- another customer acquisition channel / process to get customers scalably.

You don't want to get stuck and have to figure out that plan in the moment. You want to ramp up other channels to hedge a bit.

The question is when?

11) This is where diff marketers will disagree w/ me. Some would advise they focus on this 1 channel until they really near that saturation part. Afterall, they aren't even hitting their max CAC / max payback that they can afford to.

Others would say you need to hedge early.

12) My personal perspective is that they should focus on the 1 channel for now. And the CEO should completely hand off ALL customer acq related activities to new hires.

Once the current playbook can run completely without him, he should start experiment more w new channels.

13) What new cust acq channels should he experiment w?

My strategy is to ideally do *as similar* things as the existing playbook.

For ex: right now their process is to drive leads to a landing page & call those leads.

They should keep doing that process and try new channels.

14) So maybe it's test new ad channels. Maybe it's test more outbound emails to drive leads to that landing page.

Again, if you have a playbook that is working, try to stick with it as much as possible.

15) Examples of tactics that are very dissimilar from what they are doing:

-building self-serve!
-changing the funnel to be all inbound / free-trials

It's not to say these can't work but w/ a small pre-seed team, they need to keep as much of a working playbook the same.

16) So the new variables they will be introducing will be:

a) hiring a sales hire to take over the CEO's calls
b) being more aggressive on ad spend to hit CAC / payback limits
c) try new traffic sources to get more leads for the salesperson & monitor quality of these sources

17) They will be shelving other cust acquisition processes that are too dissimilar for now until they have a bigger team (and more VC $$).

18) The other thing about trying to keep as much the same as possible is it's easier to benchmark how small changes affect performance.

They know what great looks like. If the new salesperson can't deliver on the same playbook better than the CEO, that hire is not good.

19) In contrast, if you are changing lots of things, you don't know if issues stem from the new hire or other variables you are now introducing.

20) Often I think founders run *too many* customer acquisition playbooks at the same time. It's fine to experiment, but ultimately, concentrating on a ltd set (ideally 1 + maybe a couple of hedges) is ideal.

If you have too many playbooks at pre-seed, you can't do all well.

21) In addition, if you experiment & NONE of the channels work well because your LTV is stringently low, then that means you will be *forced* to do a lot of different customer acquisition channels in order to get a few customers here and there from many channels.

22) This is because you don't have enough "buffer" between your LTV and CAC in order to scale out a playbook / channel.

23) And, if you find yourself *needing* to spread yourself thin across many methods & channels so early, it's probably NOT a venture scale biz because the unit economics won't scale well.