What should an ecommerce growth plan include?

A plan opens by saying which part of the business is losing money and how that was worked out. Not a description of the market, not a summary of what your competitors do, and not a list of everything that could theoretically be improved.

In practice that means the four numbers behind online revenue set out for your store: how many visitors arrive, what share of them buy, what an order is worth, and how often customers come back. Then a clear statement of which of the four is weakest against your own recent history, with the drop off points named. Landing to product page. Product page to cart. Cart to checkout. Checkout to order.

The specifics are what make it a plan. Which pages, which devices, which countries, which products. A finding that says mobile conversion falls at the shipping step on your three highest traffic products can be acted on tomorrow. A finding that says your conversion rate could be better cannot.

This is the section that should contain the awkward observations, including the ones that are not about marketing. Stock that runs out on best sellers, delivery promises the courier cannot keep, a price that is out of step, a product page that does not answer the one question every review mentions.

When we write ours, this is the first page, and we build it from your public store and the numbers you choose to share. No account access is required to find most leaks, because most of them are visible to any buyer walking the store on a phone.

A plan that lists twenty opportunities is a list. A plan that says do these three, in this order, because of this, is something a team can start on Monday.

Three is a deliberate number. Most stores can genuinely run three pieces of work at once alongside trading, and doing them in sequence means you can tell which change moved which number. A plan with a dozen simultaneous actions produces a month where everything changed and nobody knows why revenue moved.

Each fix should carry four things. What it involves, in enough detail that you could brief it. Roughly how long it takes and who has to be involved, including whether a developer is needed. What number it is expected to move, named specifically rather than as growth. And why it comes before the other two, which is usually a mix of size, confidence and how much effort it takes.

Order matters more than people expect. Fixing the product feed before raising the budget, fixing the checkout before buying more traffic, and setting up the post purchase flow before chasing new customers all follow the same logic: do not pour more visitors into a container that leaks.

A good plan is also honest about the fixes that are cheap and dull. Compressing images, correcting shipping settings, removing an app that slows the page. Those rarely make an impressive slide and often produce the fastest return, so they belong in the three when they qualify.

Every plan should say what is working and should not be touched. This is the section most proposals skip, because it argues against doing more work, and it is the one that saves the most money.

Stores lose revenue to unnecessary change constantly. A redesign that removes a page layout that was converting well. A rebuild of campaigns that were profitable. Rewriting product descriptions that already rank. Replacing an email platform that was doing its job. Adding apps to a checkout that was fast.

So the plan should name the things that are earning: the campaigns with the best contribution, the pages that bring in organic visitors, the flows producing repeat orders, the products carrying the margin. And it should say plainly that these are off the table for now.

It should also name the things that look tempting and are not worth doing yet. A new channel before the current one is saturated. A loyalty programme before the basic email flows exist. A rebuild before the leak is understood. Each of those has a reason it is later rather than never, and the reason should be written down so it can be revisited.

When we write a plan, this is the third page, and it is the one clients quote back most often. Being told what not to spend money on is unusual, and it is worth more than another recommendation.

It also sets a standard for the relationship. A provider willing to say your ads are fine is a provider whose other recommendations mean something.

A plan ends with what acting on it costs and how you will both know whether it worked. Without those two, it is analysis rather than a plan.

The price should be scoped to the work described, not a general retainer for unspecified marketing. You should be able to read the fixes on the previous page and see what is included, what is not, and what the monthly figure covers. Ours is a base fee plus growth share for the scope in the plan, month to month, with no percentage of ad spend and no percentage of the sales you already make, only a capped growth share on revenue we add above your baseline. Ad spend goes to the ad platform, all of it.

Measurement should be specific to the fixes. If the first fix is a checkout change, the measure is checkout completion rate, before and after, on the same traffic. If it is a feed rebuild, the measure is impressions and clicks on the products that were fixed. Total revenue is the wrong measure for a single change, because it moves for a dozen reasons every week.

The plan should also say what happens in the first month and who does what, so nobody is waiting on anybody.

And it should be short enough to read. Four pages of specifics beats forty pages of context. What does not belong: a generic audit checklist, stock benchmarks presented as your targets, statistics that cannot be sourced, competitor teardowns, or a pitch deck about the agency.

One last test. A plan should be useful even if you never hire the people who wrote it. If it only works when they do the work, it was a proposal.