The call tends to arrive around ten weeks after launch.
The D2C site is live. The branding is sharp. The platform is performing as specified. The business case that got signed off six months earlier is still pinned to a wall somewhere in the finance director’s office. Everything above the waterline looks fine. But something has come apart underneath, and the person who was told to make D2C work is running out of runway.
What I find, when I walk into these situations, is almost always the same pattern.
Every consumer order has to be manually re-keyed into the ERP because nobody specified the integration. Two or three people in the office spend their mornings doing it, and they are behind by lunchtime. The warehouse team, which has spent decades picking pallets for trade customers, is now being asked to pick and pack single units using the same workflow. Returns are arriving back to a goods-in bay that has no consumer returns process, so they pile up next to the loading doors.
Customer service queries are landing with the team that handles dealer account admin, and that team has never had to tell a consumer their parcel is late.
The strategy deck did not describe any of this. The strategy was sound. The execution is a mess. And the gap between the two is where most D2C projects go to die.
The gap nobody owns
There is a specific space in every D2C project that sits between the strategic recommendation and the operational reality. I call it the messy middle. It is the part that involves warehouse workflows, systems integration, customer service design, pick and pack logic, returns processes, stock synchronisation, and all the other gubbings that have to work before a single consumer order can be fulfilled properly.
Consultancies are good at the front end. They map the opportunity, build the business case, and help you get board sign-off. That work has genuine value. Their engagement typically ends at the strategy. They hand over the deck, introduce you to an implementation partner, and move on.
Agencies come at the same problem from the other end. They are good at building platforms, creating content, and running campaigns. They need a brief before they start, and that brief assumes someone has already sorted out the operational questions. Most agencies do not ask those questions because they fall outside the scope of what they do.
The result is a gap in the middle of the project. The consultancy designs something the agency cannot fully implement. The agency builds something the operations team cannot run. And the business is left holding together what should have been joined up from the start.
What is actually broken
The re-keying problem, when I find it, is the visible one. It is not the real problem.
The real problem is that the ERP and the ecommerce platform have never been introduced to each other. SAP is doing what it has always done. It is holding stock data that updates once a day, in a format built for trade orders measured in pallets. The ecommerce platform is doing what it was designed to do. It is taking consumer orders in real time, expecting accurate stock to display to shoppers, and generating dispatch notes on the fly. Neither system is wrong. They just cannot speak to each other, and nobody has been asked to make them.
That single gap cascades. Stock levels on the website are out of date by the evening, which means consumers are ordering things that have already sold out, which means cancellations, which means complaints, which means customer service queries arriving at a team that has never handled consumers. The warehouse team, given no new workflow, is improvising. Improvisation in a warehouse looks like errors, delays, and costs that quietly eat the margin the consultancy promised the board.
None of this is exotic. None of it requires new thinking. It requires someone to have asked, at the strategy stage, how the two systems would talk to each other, who would pick the orders, and what would happen when something went wrong. Those questions never got asked because the people writing the strategy did not see them as strategic, and the people who would eventually build the platform were not yet in the room.
The first 90 days
This is when the sponsor starts asking difficult questions. The board wants to know why the numbers do not match the business case. The warehouse manager wants to know why consumer orders are disrupting trade fulfilment. The IT team wants to know why the integration is generating errors. And the person who got landed with making D2C work, often someone from sales or operations who had no D2C experience before this project, is trying to hold it all together.
The pattern is consistent enough that I can almost describe the state of the room before I walk into it.
What makes it worse is that by the time these problems surface, the consultancy has been paid and moved on, and the agency is busy with the next phase of the brief.
What should happen instead
The messy middle is not a phase you skip or delegate. It is the work itself.
The operational questions have to be asked at the strategy stage, not after.
How will the ERP handle single-unit orders?
What does the fulfilment workflow look like?
Who owns customer service for consumers?
How does D2C sit alongside existing trade relationships?
What does the returns process look like, physically, in the warehouse?
Where does the product data come from, and who is responsible for it?
These are not exciting questions. They do not make for a compelling slide deck. They are the questions that determine whether a D2C channel actually works or becomes an expensive embarrassment.
Beyond asking them, somebody has to own them through delivery. Not as a strategic advisor who attends the weekly meeting. As a person in the room who worries about the details. Whether the pick and pack logic is working. Whether the stock levels are syncing. Whether the returns process makes sense when a consumer actually tries to use it. Whether the customer service scripts fit the tone the brand has built everywhere else.
That is the difference between D2C that works and D2C that does not. Not the strategy. Not the platform. The presence of somebody who cares about the gubbings.
By the time I get called in, that person is always hired late. The first thing we end up doing is spending three months fixing things that should never have broken.
