Outbound vs inbound for B2B: which constraint do you have
Outbound and inbound solve different constraints, not the same one at different speeds. How to tell which you have, and the legal floor under outbound in the Gulf.
Read
The argument is usually had as though it were about the number. It is not. It is about who carries the risk of the work taking longer than anyone thought, and the two models answer that question in opposite ways.
Ask a services business whether it prefers retainers or project fees and you will get a preference dressed as a principle. Retainers are predictable, projects are honest, and everybody has a story that proves their side. The stories are all true and none of them are the reason.
The useful way to separate the two is to ask what the client is buying. A project fee buys an outcome with a defined edge. A retainer buys ongoing access to capacity. Those are different products, and most of the disputes that end a client relationship come from selling one and delivering the other.
A project price is a bet that you understand the scope. If you are right, the margin is yours. If you are wrong, the overrun is yours too, and the client is insulated from it by design. That is not a flaw in fixed pricing, it is what the client paid for.
A retainer moves that bet. The client buys a month of your attention and carries the risk that the month is quiet or that the work turns out harder than expected. In exchange they get responsiveness they cannot get from a vendor who has to quote before starting.
Neither is more professional than the other. What is unprofessional is charging a retainer and then behaving as though every request needs a scoping conversation, or charging a project fee and treating the scope as advisory.
This is not only a commercial distinction, it is a reporting one, and the standard is unusually clarifying about what the difference really is. IFRS 15 sets out that an entity recognises revenue "to depict the transfer of promised goods or services to the customer in an amount that reflects the consideration to which the entity expects to be entitled in exchange for those goods or services".
The step that matters here is identifying the performance obligation. ACCA's summary of the five steps describes a performance obligation as "a distinct promise to transfer specific goods or services, distinct from other goods or services", and sets out when revenue is recognised over time rather than at a point in time, including the condition that "the vendor has an enforceable right to be paid for work completed to date". That guidance is in ACCA's technical article on the IFRS 15 revenue recognition steps.
Read that as a commercial test rather than an accounting one and it does real work. If your retainer is a series of substantially similar services delivered month after month, you are selling access, and the revenue arrives as the months pass. If your engagement is one deliverable that only has value when it is finished, you are selling an outcome, and pretending otherwise creates a contract where you can do eleven months of work and be owed nothing.
Both sides feel this and they feel it differently, which is why the conversation gets tense without either party being unreasonable.
| Dimension | Retainer | Project fee |
|---|---|---|
| Who carries scope risk | The client | The vendor |
| Cash timing for the vendor | Even, and forecastable | Lumpy, tied to milestones |
| Cost visibility for the client | Known monthly, unknown per output | Known per output, unknown in total over a year |
| What a quiet month means | Vendor profit, client resentment | Nothing, the fee was for the outcome |
| What a hard month means | Vendor loss, client relief | Vendor loss, invisible to the client |
| How it usually ends badly | Client asks what they are paying for | Vendor stops answering after delivery |
The two rows in the middle are the ones worth reading twice. A retainer and a project fee fail in mirror image. The retainer fails on a quiet month, when the client cannot see what they bought. The project fails after delivery, when the vendor has no commercial reason to still be present and the client discovers that nobody owns the thing now.
Pricing structures teach behaviour, and they teach it faster than any statement of values in a proposal.
A fixed project fee rewards finishing. It also rewards defending the scope line, because every additional request is unpaid work, and a vendor under margin pressure becomes a vendor who says no politely and often. Clients read that as rigidity. It is arithmetic.
A retainer rewards being useful continuously, which is a better incentive, until it becomes a reward for looking busy. The failure mode is a monthly report that documents activity rather than results, which is the same disease we described in the six numbers that earn a place on a dashboard. If the retainer's evidence of value is a list of things done, the retainer is being defended rather than earned.
A retainer is the right structure when the work genuinely recurs, when responsiveness has value in itself, and when the client would otherwise be paying you to re-learn their business every time they call. Ongoing pipeline work fits this well, which is why business development is usually retained rather than bought by the project, a point we made in outsourced BD versus hiring a salesperson.
It is also the right structure when the client needs someone to hold context. A consultancy that has to be briefed from zero on every engagement is expensive in a way that never appears on the invoice.
Take a project fee when the deliverable has a real edge, when you have done something close enough before to price it, and when the client can genuinely say what finished looks like. If they cannot, a fixed price is not protecting them. It is converting an unclear brief into a dispute with a date on it.
The same logic appears in software, where a defined deliverable with staged payment is a well understood structure. Rivl set out how that works in practice in pay on delivery software development, and the mechanism transfers cleanly to any service where the output can be inspected.
Most mature service relationships end up in the same shape, and it is worth going there deliberately rather than after an argument. A modest retainer covers the continuing work, the context holding and the responsiveness. Discrete pieces with real edges are quoted separately as projects.
The one rule that makes it hold is that the boundary is written down before it is tested. Name what the retainer includes, name the threshold at which something becomes a project, and agree it while everybody is happy. A boundary negotiated during the first disagreement is not a boundary, it is a concession.
None of this fixes an underpriced engagement. If the number is wrong, both models lose money and they only differ in how long it takes to notice. Working out the number is a separate exercise, and we set out a method for it in how to price a B2B service without guessing.
There is also a size below which this whole question is theoretical. If you have one or two clients, the structure that matters is the one they will sign, and optimising the model before you have enough work to have a pattern is a way of avoiding selling. Get to a handful of engagements first, then look at which of them are quietly losing money and why. The answer is often the structure rather than the price.
Send us your last three engagements with what was agreed and what actually got delivered. We will tell you which ones are mispriced and which ones are misstructured, because they are different problems with different fixes.
Book a Free Consultation