Your eCommerce Launch Was Decided Ninety Days Before It Launched

The Part of the Plan That Is Already Late

Ask a company when its eCommerce channel launches and you get a date. Ask what happens in the ninety days before that date and the answer tends to be about the site: catalog loading, integrations, design review, QA.

That is phase one work. It is real, and it is not where these launches are usually won or lost.

The pattern we keep meeting is a business that has funded a great digital experience and has not asked whether the commercial side of the house can serve the demand that experience creates. The site goes live, the orders arrive, and the company discovers that the SKUs chosen for the channel were chosen by whoever built the catalog, that inventory was planned against last year’s wholesale forecast, that the service team is answering two-day delivery questions with two-week answers, and that the warehouse is picking eCommerce orders in the same lane as pallet shipments.

None of that is a website problem. All of it shows up as one, because the site is where the customer meets it.

So the honest version of a launch plan starts earlier than most plans do. Call it phase zero, running from roughly ninety days out to a month out, and treat it as the phase with the power to stop the others.

Phase Zero Is Four Conversations, None of Them Technical

Merchandising. Which styles and SKUs go online, and why those. The reason this needs an explicit answer is that the default answer is everything, and everything is rarely the right assortment for a channel with different economics, different shipping realities and a different buyer. Demand planning has to cover both the existing customers who will migrate and the new customers the channel is meant to attract, which are different curves. Pricing and margin need a position across channels before launch rather than after the first complaint. And where a vendor can already ship direct faster than you can, drop ship is worth activating rather than absorbing all the fulfillment yourself.

Inventory. Vendor commitments tied to the forecast, and velocity modeled at SKU level against predicted channel demand rather than against total demand. Allocation rules matter here more than people expect, because marketplace and direct channels are not interchangeable: a marketplace fulfillment program can carry strict inventory requirements that a direct channel does not, and planning them as one pool tends to starve whichever one is treated as the overflow. If drop ship is part of the plan, the integration and the returns path need testing before launch, not during it.

Customer service. The expectation gap is the thing to prepare for. A long-standing B2B customer may accept a two-week lead time as normal. The same person buying through the channel expects days, and expects to be told where the order is without asking. That is a training and staffing question ahead of launch, plus protocols for returns, exchanges and escalations that the team has actually rehearsed.

Fulfillment and operations. Direct orders behave differently from wholesale orders: smaller, more frequent, individually packed, individually tracked. That usually means separate procedures rather than adapted ones, often a dedicated picking and packing area, and a test run at something like real volume before anyone is depending on it. The post-purchase experience belongs in this conversation too, because confirmations and tracking are operations dressed as marketing.

Underneath all four sits the unglamorous one: a named owner per stream, a shared plan, and a standing cross-functional check-in. The launches that struggle are rarely short of effort. They tend to be short of anyone whose job it was to notice that two streams had drifted apart.

The Gate That Makes It Real

A phase with no gate is a suggestion. The version we would defend puts a hard checkpoint at roughly T minus forty-five, before any commitment to the launch date is confirmed publicly or internally.

The gate asks whether the assortment and inventory plans are complete and loaded, whether drop ship flows have been tested end to end, whether a fulfillment run has simulated peak load successfully, whether service is trained with its tooling live, and whether every function has signed its own readiness rather than had it signed for them.

If the answer to any of those is no, the useful move is to move the date. That is unpopular and it is far cheaper than the alternative, because a launch into unready operations does not fail quietly. It produces a cohort of first customers who experienced the worst version of the channel, and those customers are both the hardest to win back and the most likely to talk.

A second gate belongs around two weeks out, covering the digital work: catalog quality, integration stability, the first campaigns built, journeys consistent across web, mobile and email, analytics actually firing.

Launch Narrow, On Purpose

When the channel does open, the instinct is to open it fully. We would open it deliberately narrow.

The launch assortment should be the SKUs that are best positioned to convert and that operations can certainly serve, chosen on margin, fulfillment readiness and fit rather than on breadth. The goal in the first month is not range. It is a channel that works, observed under real load, with conversion, fulfillment service levels, customer satisfaction and returns watched daily rather than monthly.

Then a third gate at roughly two weeks after launch, before scaling spend: are the first campaigns clearing their return threshold, is service holding its standard, are the conversion bottlenecks identified rather than guessed at, and is fulfillment stable under actual volume. Scaling acquisition spend into an unstable channel is the most reliable way to turn a small operational problem into an expensive one.

From there the work changes character. Expanding assortment on evidence of sell-through, improving site performance and search on real behavior, moving analytics from reporting what happened to modeling customer value, and only then considering marketplace expansion, on the condition that inventory and service could absorb several times the current volume without rework.

Four Questions Before You Fund It

If you are a sponsor looking at an eCommerce launch inside a value creation plan, four questions will tell you more than the roadmap will.

What happens in the ninety days before launch, and which of it is not about the website?

Who has signed that operations, inventory, merchandising and service are ready, by name?

What is the gate that can move the launch date, and who is allowed to pull it?

What is the launch assortment, and why those SKUs rather than the whole catalog?

The launch date is the least informative number in the plan. What matters is whether the ninety days in front of it were spent on the site or on the business that has to stand behind it.

Like what you are reading? Let’s talk.

If this reflects the kind of work you need, it might be time to bring in a leader who has done it before.