Use Feedback From a Lost Commercial Bid Without Inventing Why You Lost
Use Feedback From a Lost Commercial Bid Without Inventing Why You Lost
The rejection call takes four minutes. The email that follows is shorter still — a thank-you, a decision, and two or three sentences about the work. Then your team spends an afternoon producing a much longer document about why it happened, and almost none of that document is in the email.
That afternoon is the problem worth fixing. There are two jobs in front of you, and only one of them has an answer sitting in the room. The first is establishing exactly what the recipient said, about which passage, on which version they saw. That's a matter of record. The second is deciding whether any of it changes what you'd do next — and that one stays a judgment, because you still can't see the selection.
So the review has a shape. Separate what you were told from what you concluded. Put the comment next to the passage it describes. Compare one real alternative rather than a slogan. Ask a bounded question if the answer would actually change your reading. Then carry forward the smallest change you can justify.
Nobody involved in writing this watched your pitch or anyone else's, and no industry data is being reported here. This is a method for working with partial information, not an account of how commercial selection works.
The example used below is invented. The brand, brief, treatment and client comment are constructed to show the method. No real pitch, client, winning studio or selection decision is represented.
Halden is a fictional bicycle maker with a fictional folding commuter bike. The brief asked for a 45-second film, no dialogue, for a trade-show loop and the brand's own channels, and it contained this line: The fold is the product's defining action. The film must make clear what the object is and what it becomes. The team submitted a treatment, then a revised treatment with boards. The comment arrived in writing, from the brand's marketing lead, after the decision was communicated — and it responded to the second submission.
Separate the result from the information received
Three kinds of material arrive in the same week and they should not live in the same place.
Recorded. The outcome, the date, and which version the recipient saw. In the Halden case: the decision went to another studio, the comment came after that call, and it referred to the second submission. That much is known and should be written plainly.
Reported. The recipient's actual words. Not your summary of them, not the version you'd have written if you'd been drafting their email. Copy them in.
Interpreted. What the team believes the words mean and what caused the loss. This column is where the danger lives, because it fills up fastest and feels most like fact by Thursday.
Keeping these apart is not pedantry. "We lost because the treatment was too atmospheric" is a claim about a room you weren't in. "The marketing lead wrote that she lost track of the object before it folded" is a claim about a sentence. Both may end up in your notes. Only one of them can be checked.
Watch for three specific conversions, because they happen almost automatically:
Politeness becomes approval. "The fold itself is lovely" is not a statement that the film worked. It's often a way of earning the right to say the harder thing in the next sentence, and it's frequently the most social part of an email that has to be written at all.
Silence becomes a defect. If the recipient said nothing about your casting, your music or your length, that is not evidence about any of the three. It's an absence, and absences are cheap to fill with whatever the team already suspected.
Absence of explanation becomes evidence against you. A short rejection usually means a busy person wrote a short email, not that the work failed in the way you're most afraid it did.
None of this means the comment is worthless. It means the comment is evidence about the work, and the decision is evidence about the decision. Keep them apart on the page and the review gets shorter and more honest at the same time.
Put the comment beside its passage
A comment earns its value when you can locate it. Before you decide anything, find the material it describes and ask what the brief required that material to do.
Here is the invented passage the Halden comment described:
0:00–0:18 Rain. Three intersections, close shots only: gloved hands on the bar, a wheel through standing water, a grey coat, water on the lens. The bicycle is never seen whole.
0:18–0:26 At the station entrance she swings the frame; the bike folds in one continuous motion; she carries it through the turnstile.
0:26–0:45 Train interior. The folded bike under the seat. She reads. End card.
And the invented comment:
The fold itself is lovely. But by the time she folded it I had lost track of what the object was, and at the turnstile I wasn't sure the thing she was carrying was the same thing she'd been riding. We were also hoping for something with a bit more warmth.
Two sentences, two very different kinds of information. The first names a place in the film: the opening eighteen seconds and the fold. The second — more warmth — describes a whole piece that doesn't exist, and there's no passage you can hold it against. Record both. Act on one, and note that the other can't yet be diagnosed.
The locatable comment doesn't require you to guess at taste. It points at sequencing. The film's first act shows the bike the way you'd shoot a suspect: hands, spokes, coat, rain, no whole object. Then at 0:18 the bike folds, and the folded thing is, visually, a new object rather than a change of state in one we already know. The brief asked the film to make clear what the object is and what it becomes. The passage is strong on what it becomes and thin on what it is.
That's a local defect you can see in the material itself. Whether it had anything to do with the decision is a separate question, and no amount of staring at the treatment will answer it. A local problem can be obvious while its importance to the outcome stays unknown. Both facts are true at once, and the useful review holds them together instead of picking the one that feels like closure.
The team's own explanation in this invented case was different: the film led with mood instead of product, they said, and that's why it lost. Notice what that claim requires — knowledge of what the other submissions did, how the brand's stakeholders weighed a trade-show loop against a channel film, whether budget was a factor, what the internal conversation was. The team saw none of it. The explanation is not a finding. It's a story that got told first because it was available.
Compare a real alternative, not a hindsight slogan
"Next time we'll make the product clearer" is a slogan. It's unfalsifiable, it can't be reviewed, and it will be agreed to in every room you bring it into. A real comparison means rewriting the affected passage — hypothetically, on paper — and stating what the change gains and what it costs.
Here is the same invented film with one bounded revision:
0:00–0:05 Rain. Two intersections, close shots as before.
0:05–0:07 A medium shot in profile: the rider passes a lit shopfront and the whole bicycle is readable for two seconds — frame, wheels, handlebar — before the tight shots resume.
0:07–0:18 Back to close shots.
0:18–0:26 Unchanged. She swings the frame, the bike folds in one continuous motion, she carries it through the turnstile.
0:26–0:45 Unchanged.
What becomes clearer: the folded object now has an antecedent. The fold reads as a change of state rather than an appearance, and the turnstile moment stops being a question. The brief's second requirement is served without touching a frame of the fold itself.
What the change gives up: two seconds out of the anonymous opening. The tight-shots-only rule of the first act was doing something — it kept the rider a stranger and made the city the subject. Inserting a readable object at 0:05 announces the film's subject earlier and weakens that reading a little. It's a real cost, and it's smaller than the cost of the current sequence, which is why the trade is worth making here.
And this is a paper comparison. Nothing was shot, tested or shown to the client. A drawing of a two-second shot is not a screened two-second shot; it can teach the sequencing method without pretending to be a result.
Two cautions apply whenever you do this.
Don't convict the whole submission. The rain was not on trial. Neither was the grey coat, the silent film, or the choice to end on the train. When a bid is lost, every feature of the submission starts to look like a suspect, and teams often strip the strongest, most deliberate choices because those are the easiest to blame. The question for each feature is not "did we lose?" but "did this serve the brief as written?" If it did, and the comment didn't touch it, it survives the review. In the invented Halden case, the fold stays uncut.
Stay inside one alternative. Two or three branches at once and you're back to opinion, because nothing can be compared. Rewrite one passage, state the gain, state the cost, and leave it there.
Ask a bounded follow-up when it could change the lesson
Some comments have two readings, and the difference matters. In the invented Halden email, "I had lost track of what the object was" could mean the sequence genuinely failed to establish the bike, or it could mean the fold was understood perfectly and the impression of vagueness is how a different objection got phrased — that the film felt slow, or that the product's role in the film wasn't what this stakeholder wanted. Those lead to different next moves. One is a sequencing fix. The other isn't a fix at all.
A follow-up is worth asking when the answer would actually separate those readings. It should be short, specific, and answerable in a line or two by someone who has minutes, not an afternoon:
One question if you have time, and no problem at all if you'd rather not: in the middle section, was it clear that the folded object was the bicycle from the opening shots, or did the film lose you earlier than that?
That question asks about the film, not the competition. It's the kind of thing a recipient can answer without exposing anything.
What not to ask: which studio won, what the winning treatment did, how you scored, what the budget comparison looked like, what was said in the room. Those requests put the recipient in an awkward position, tell you less than you think, and cost you the relationship you'll need on the next brief. If a private debrief happens, it's because the recipient offered it.
Route the question through whoever you already deal with. Ask once. If the answer comes, treat it as one person's account — useful, attributed, not proof that this is why the work lost. A brand lead is often describing a decision they were one voice in.
If no answer comes, the diagnosis simply stays bounded. That's a legitimate result of the review, not a gap you're entitled to close. The rule that follows from it is worth saying out loud in the meeting: no reply does not authorize the team to invent the answer.
Carry forward the smallest warranted change
The likely output of this review is one adjustment, one retained choice, and one open question.
For the invented Halden case, the adjustment is a check, not a house style: before any product changes state on screen, show it whole first — unless the brief asks us to withhold it, in which case withhold it on purpose and say so in the treatment. That formulation matters. A check that includes its own exception is a piece of judgment. A rule without one hardens into sameness across every film, and every client gets the same opening beat.
The follow-through is small. In the next read of a treatment, find the frame where the product enters and ask what the audience knows at that moment. If the answer is "fragments," either add the whole view or confirm that the withholding is deliberate. One clause in the review, applied to one passage.
Retained without apology: the fold in one continuous motion, the rain, no voiceover, no card explaining the object. The comment praised the fold. Nothing in the email touched the rest.
And record the basis beside the uncertainty, together, so the next person who reads the file doesn't inherit a conclusion you never reached:
Halden — invented example. Observed: written comment on the second submission, from the brand's marketing lead, that the product's identity was unclear before the fold, plus a general note about warmth that named no passage. Proposed adjustment: show the product whole before its state change where the brief requires the change to land. Retained: the uncut fold, the rain, the refusal to explain the object in words. Cause of loss: unknown. We did not see the other submissions, the scoring, or the room.
One lost bid is one data point about one assignment. It does not establish that length causes losses, that atmospheric openings cause losses, or that a particular collaborator does. It's entirely possible the next brief rewards exactly what this one didn't — and if you've rewritten your method around an invented cause, you'll have removed the thing that was working before you ever find out.
Frequently asked questions
What should I separate when reviewing a lost bid?
Separate recorded facts, reported words, and interpreted conclusions. Recorded includes outcome, date, and which version the recipient saw. Reported is the recipient's actual wording. Interpreted is what the team believes the words mean and what caused the loss.
What should I do with feedback that does not point to a specific passage?
Record it, but note that it cannot yet be diagnosed or acted on. A comment like a request for more warmth describes a whole piece that does not exist and has no passage to hold it against. Act on the comments you can locate.
How do I avoid inventing a cause for the loss?
Do not claim why the work lost when you did not see the other submissions, scoring, room, or selection process. A local problem can be obvious while its importance to the outcome remains unknown. No reply does not authorize the team to invent the answer.
What makes a real alternative instead of a hindsight slogan?
Rewrite the affected passage hypothetically on paper, state what the change gains and what it costs, and stay inside one alternative. Do not convict the whole submission. A slogan like make the product clearer is unfalsifiable and cannot be reviewed.
When is a bounded follow-up question worth asking?
Ask when the answer would separate two readings and could be answered in a line or two. Ask about the work, not the competition, and route it through whoever you already deal with. If no answer comes, the diagnosis stays bounded.