First 90 days of a marketing build, month by month

Aug 22
4 min read

Updated: Aug 27

Introduction

A fixed-scope engagement has an advantage over an open-ended one: it has to finish. That constraint forces decisions about sequence that a retainer can postpone indefinitely.

Ninety days is enough to install something that works and to train the people who will run it. It is not enough to do everything, which is why the ordering matters more than the ambition.

1. The first 90 days of a marketing build start with measurement, not activity

Weeks one and two produce no visible marketing and determine whether the rest is worth anything.

What gets established: where enquiries currently come from, what a customer is worth, what the conversion rate is from enquiry to sale, and what is being spent. Plus account ownership — every platform registered to a business-controlled address.

Businesses often resist this phase because it looks like preparation rather than progress. Skipping it means every later decision is a guess, and there is no way to demonstrate what the engagement achieved.

2. Fix the conversion points before anything is promoted

Weeks three and four. Traffic multiplies whatever conversion rate already exists, including a poor one.

The work is unglamorous: the form tested on a phone, the response time to enquiries, the phone answered, the price visible, the booking route working, the listing complete.

This is also where the fastest results appear. Businesses frequently see enquiry volume rise in month one without any increase in spend, purely from stopping the leaks.

3. Sort the offer and pricing in month one

Still within the first thirty days, because it costs nothing and changes everything downstream.

What is being sold, at what price, in what structure. Margin per item or per hour. A defensible price list, one add-on, and a reason to choose the middle option.

Advertising an unclear or badly priced offer buys more people to lose slowly. Pricing is the only lever that reports back immediately, which makes it the right work for the period before any channel is running.

4. Install one channel properly in month two

Days thirty to sixty: acquisition, and only now.

One channel, set up correctly with conversion tracking, negative keywords, a landing page matched to intent, and a capped test budget. Not four channels launched simultaneously.

One channel run properly produces a decision. Four run partially produce activity nobody can evaluate, which is the most common way a build's budget disappears without a finding attached.

5. Build the retention routine alongside it

Also month two, because it is cheap and it changes the economics of the channel.

A welcome sequence, a review request that fires reliably, a reason to return, and a way to identify customers who have gone quiet relative to their own pattern.

This is what makes acquisition affordable. A business where customers return can pay more to acquire them, so retention work expands what the channel can do rather than competing with it.

6. Month three is documentation and training

Days sixty to ninety, and the phase most often compressed into a final week.

Every recurring routine written down: what it is, who owns it as a role, when it happens, what it produces. The weekly check, the monthly review, the maintenance tasks. Then the team is trained by doing it, not by watching.

This is the difference between buying activity and buying an asset. A system nobody internally can operate reverts to nothing within two quarters of the engagement ending.

7. Hand over ownership explicitly, not gradually

The final week is a transfer, and it should be a defined event.

Administrator access on every account confirmed from a business address, all data exported and verified, editable source files delivered, and a written document explaining what is running and why — including what was tried and abandoned.

The reasoning is the part that cannot be reconstructed later. Configuration is visible; the thinking behind it is not.

8. Set the review date before the engagement ends

Ninety days installs a system. It does not prove the system works, because some channels need longer than that to be judged.

So the final deliverable includes what to check and when: which numbers, at what dates, and what result would justify changing course. Leading indicators for anything slow.

Also state honestly what has not been proven yet. A build that reports certainty about a channel tested for four weeks is overstating what ninety days can establish, and the client will discover that on their own.

Conclusion

Measure and secure account ownership first, fix the conversion points and the offer inside month one, install one channel properly in month two alongside a retention routine, and spend month three documenting the routines and training the people who will run them.

Close with an explicit transfer of accounts, data, files and reasoning, plus a written review schedule and an honest statement of what has not yet been proven.