What is a good ecommerce marketing budget split?

Arguments about how much should go to search versus social miss the point, because the same platform can do two completely different jobs. A budget split only makes sense when it is organised by what the money is being asked to achieve.

There are three jobs. Capturing demand means putting yourself in front of people who are already looking: brand searches, category searches, shopping listings, retargeting the people who visited last week. Creating demand means reaching people who were not shopping and giving them a reason to start. Keeping customers means everything that produces a second and third order from someone you already paid for.

Once the money is grouped that way, the split becomes answerable, because each job has a natural size. Capture is limited by how much demand exists. Create is limited by margin and patience. Keep is limited mostly by how much work you put in rather than how much you spend.

This also stops a common mistake. When capture and create sit in the same reported number, the blended return looks fine while the demand creation part quietly fails, or the demand creation part gets cut because it looks expensive next to campaigns that are simply collecting people who already knew you.

Write the three groups on the plan, put every campaign in exactly one of them, and report each separately. The split then becomes a decision you can defend instead of a habit.

Capture is the cheapest money you will spend, so it gets funded before anything else and it gets funded to its ceiling rather than to a percentage.

The ceiling is real and it is set by demand. There are only so many people searching for your brand, your products and your category this month, and once you are showing for most of those searches, more budget buys worse matches. The signal is straightforward: impression share is high, and adding budget raises cost per click while orders stay flat.

So the way to size capture is to increase it in steps and watch the cost per order at each step. Stop where the extra spend stops producing extra orders at a price you accept. That number is your capture budget, and it changes with season, catalogue size and stock.

Inside capture, put brand and high intent product searches first, then category terms and shopping listings, then retargeting. Cap retargeting deliberately, because it is very easy to spend a lot showing ads to people who were going to buy anyway. If retargeting is a large share of the budget and the total number of new customers is flat, you are paying to take credit for orders you already had.

One caution. A store with almost no brand demand has a small capture ceiling, and topping up the budget there will not help. That store's problem is that nobody is looking for it yet, which makes the next section the important one.

Whatever the business can afford beyond capture belongs to demand creation, because that is the only part of the budget that grows the number of people who know you exist.

Judge it on different numbers. Cost per new customer, not blended return. Total revenue against total marketing cost across the whole business, so the effect on brand searches and direct visits is counted somewhere. And a longer window, because a video someone watches this week may produce a search next month.

Budget the production as part of it. Demand creation runs on creative, and creative wears out. If the media budget is set with nothing left for photography, video and editing, the campaigns will run out of things to say and the cost per new customer will climb. A working rule is to fund creative production as a named line rather than an afterthought.

Size it against ambition and patience. A store that wants to hold its position can put a small share here. A store that wants to grow meaningfully has to accept a period where reported return falls while the customer base widens, and the honest way to make that decision is to know the repeat rate. If a first order reliably produces a second, you can afford to buy customers at close to break even. If it does not, demand creation has to pay for itself immediately and the budget stays small.

Set the split, then review it quarterly rather than annually, because demand and costs both move.

The third job usually needs the smallest share of money and produces the largest share of profit, which makes it the part most often under resourced.

Email and text cost very little to send. What they cost is the work: writing the flows, building the segments, producing the content, keeping the list healthy, and setting up the subscription or replenishment mechanics. So the retention line in a budget is mostly a people line, and the mistake is to judge it against a media spend.

Because the money is small, the way to size it is by outcome. Decide which flows exist by the end of the quarter, which segments get their own message, and what share of revenue should come from returning customers. Then resource that.

There is a media component and it is worth funding: retention advertising to existing customers about a new product, and audiences built from your list. Keep it separate from prospecting so it does not flatter the new customer numbers.

One more line belongs in every budget: a reserve for testing and for seasonality. Testing means a fixed share held back for new channels, new creative angles and new offers, spent whether or not the month was good, because a store that only spends on what already works finds out too late that it stopped working. Seasonality means the split is not the same in a peak month as in a quiet one. Move money towards capture when demand spikes, and towards creation and retention when it does not.