The first decision without a marketer is the cost ceiling per customer

At a company with no dedicated marketing person, once you start looking into how to get customers, it goes like this.

You should post on Instagram every day. You should write a blog too. You should tidy up the Google profile too. You should collect reviews too. You should try ads too. All of that is right, and none of it is wrong.

That is why it does not get decided.

You can make an infinite list of things to do. The problem is not that the number is large. It is that there is no criterion to line them up with. Once you decide 1 criterion, the options shrink on their own.

I am building Shoryaku AI ( https://shoryaku.net ), a marketing-support SaaS. This article is about how to set that criterion, and the order after you set it.

3 common failures

If you proceed with no criterion, you usually get stuck in the same places.

The 1st is making something free and quickly finished the top priority.

"This is free, and it can be done in 30 minutes, so start here." The order is decided by how easy it is to start. Something cheap and fast, and something that affects revenue, are completely different matters.

The 2nd is being satisfied once you have collected the numbers.

You compare the admin screens, put them in a table, and stop there. Even when the numbers are lined up, what to do next is not written down.

The 3rd is starting 3 things at the same time.

The more motivated you are, the more this happens. Then, 1 month later, you no longer know what worked. You keep the thing that was not working, and you stop the thing that was. That is the judgment you end up making.

All 3 of these can be avoided if you have a criterion for deciding the order.

The first thing to decide is how much you can spend to acquire 1 customer

This is the 1 thing to decide. How much is it acceptable to pay to acquire 1 customer.

Call this the CAC ceiling. The calculation is 2 lines.

LTV         = revenue per 1 customer × gross margin × the number of times it continues
CAC ceiling = LTV ÷ 3

"The number of times it continues" is the number of months it continues for a monthly service, and for a visit-based business it is visits per year times how many years. It does not have to be exact. If you do not know, it is fine to place it by feel. Placing it always makes the judgment better than not placing it.

Even for the same product at ¥4,980, the answer changes by more than double.

The business above can still recover even if it spends ¥20,000 to acquire 1 customer, but the business below goes into the red if it does the same thing.

Once this number is out, half of the options disappear on their own. Suppose you were considering an ad at ¥500 per click. If it takes 100 clicks to reach an application, that is ¥50,000 per customer. For a business whose ceiling is ¥9,300, it drops out of the candidates at that point. You can see it on the desk, before you try it.

And failure 1 from the opening disappears. Things line up by "does it fit inside the ceiling," not by "it is free, so it is the top priority".

4 steps to take after you decide

Step 1: Make the current numbers visible

The ceiling is decided, but you cannot choose a move if you do not know where people are dropping off now.

Google Analytics, Search Console, and Google Ads. These 3 can be put on 1 page for free in Looker Studio. Do this before you consider a paid tool.

Do not add outcome counts across channels. Google Ads conversions and GA4 key events are counted differently, so if you total them you get a number larger than the real one. Only line them up and compare them (more detail is here).

Step 2: Decide the 1 place that is stuck

Getting customers splits into these 4 stages.

New visits → conversion on the site → inquiry or booking → retention or repeat purchase

Decide only 1 place, the one that is the thinnest.

The way to tell is simple. If people are coming to the site and there are no inquiries, what is stuck is not the inflow. Adding ads here only adds water to a bucket with a hole in it.

Increasing acquisition when the inflow is already enough. I think this is the most common mistake. Failure 2 from the opening happens when you look at the numbers without deciding this.

Step 3: Split the moves by whether they disappear when you stop, or remain

Split the candidates into 2 kinds.

Do not line these 2 up on the same "what to do this month" list. If you do, the one that works sooner always comes out on top.

Then see whether there is a paid move that fits inside the CAC ceiling. If there is not, hold the ads and start from the cumulative-asset kind. If you do run ads, start by finding keywords where competitor bids are thin.

Step 4: Carry out only 1, and look at 14 days and 28 days later

Do not start 3 at the same time. This is failure 3 from the opening.

Pick 1, decide the day you will run it, and put on the calendar a plan to look at the same numbers 14 days later and 28 days later. If you do not decide the period, you judge whether it was good or bad by how you feel.

And even when a result comes out, stop at "it improved," and do not assert "this was the cause". In the same period, the season and the competitors are moving too. What you can say stops at "an improvement was observed". If you assert it here, the next time you do the same thing and miss.

What to do in the first week

Turned into work, the above is 5 things.

1. Produce the LTV and the CAC ceiling (30 minutes). The gross margin and the number of times it continues can both be guesses
2. Put the Google numbers on 1 page in Looker Studio (1 hour)
3. Of the 4 stages, decide the 1 that is the thinnest
4. Pick only 1 move that works on that stage. Drop anything that exceeds the CAC ceiling
5. Put the run date, and the check dates 14 days and 28 days later, on the calendar

All of this can be done without buying a tool. I think you should buy what is missing after you have done this far.

When running this every month is hard

Shoryaku AI is a service that turns these 4 steps into 1 flow, as they are.

When you register the business information, the LTV and the CAC ceiling are calculated automatically (an item left blank is filled with the default for the industry, and it is marked as an estimate). It gathers the numbers from the connected channels, judges where it is stuck, and then puts out this week's moves with the evidence, the steps, and the KPI. A move that exceeds the CAC ceiling is dropped on the spot.

Whether it is a cumulative asset or a flow, and about how long until an effect appears, are also shown for each move (how it works is here). Moves you have run are compared at 14 days and at 28 days, and cause is not asserted there either.

What it supports is the 5 of Google Analytics, Search Console, Google Ads, Meta Ads, and Instagram. It goes as far as the proposal, and it does not operate the ads.

To try it

The 14-day free trial can be tried with no card registration at the start. How to choose a plan is in another article.

If you want to look at only your own numbers first, use the CSV sales diagnosis, which does not require signup.

• Marketing
• AI
• Product