Testimonial Video Editing
the interview questions decide the edit before you touch a timeline

Ask a sequence, not a list: what was the situation before, what specifically made you hesitant to buy or switch, what tipped the decision, what specific result have you seen since, and would you recommend this and to whom. This structure produces a natural narrative arc (before, objection, decision, after) that maps directly onto an edit, rather than a scattered set of positive but disconnected statements.

Ask for numbers explicitly and more than once, phrased differently each time — 'can you put a number on that', 'what would you estimate that saved you' — because customers rarely volunteer specific metrics unprompted, even when they know them, and a specific number is consistently the single strongest element in the final cut.

A sentence earns a place in the edit if it contains a specific result, names the objection that was overcome, or describes a concrete before/after state. Generic sentiment ('the team was great', 'we love the product') should almost never appear as a standalone statement in the final cut — it can function as connective tissue between specific statements but should never be the thing being highlighted.

Prioritise moments where the customer's phrasing sounds unprompted and natural over moments that sound like a rehearsed pitch, even if the rehearsed version is more polished — audiences are calibrated to detect over-coached testimonials, and a less polished but clearly genuine statement earns more trust.

Structure the cut as: specific result stated as the opening hook, brief context on the prior situation, the objection or hesitation named honestly, the decision point, then a close on the recommendation. Opening on the result rather than building up to it front-loads the payoff for viewers who won't watch to the end — a pattern that consistently improves performance for this format.

Cut in b-roll of the actual product in use, the customer's workplace, or relevant context at a minimum 40:60 ratio to talking-head footage; a testimonial that is 100% static talking head for 90 seconds loses viewers regardless of how good the content is. Avoid stock b-roll entirely for this format — generic office stock footage undermines the specificity and authenticity that makes a testimonial worth producing in the first place.

Every testimonial needs written customer approval of the final cut before any public use, not just verbal agreement during the interview — customers can and do change their minds about specific phrasing once they see it published, and a written approval trail protects both the customer relationship and the company using the footage. Send the final cut with a clear approval deadline and an easy way to flag specific timestamp concerns.

If a customer requests a specific figure be removed or softened after seeing the cut, honour it even if it weakens the video — a customer relationship damaged by a video used against their wishes costs far more than one slightly less punchy testimonial.

Before recording, get the specific result from the account or customer success team — do not rely on the customer to remember or volunteer it cold on camera. Structure the interview using the situation-objection-decision-result-recommendation sequence above, and ask every question twice in different phrasing if the first answer is vague, since a slightly awkward second attempt at a question usually produces the specific detail the first pass missed.

In the edit, find the single sentence containing the clearest specific result and put it first, even if it was said last in the interview — testimonial edits are rarely chronological. Layer in b-roll from the customer's actual environment or product usage, keep the total length under two minutes, and send it for written sign-off before it goes anywhere near a sales deck or ad account.

Asking open-ended, generic questions ('tell me about your experience') that produce polite but unusable general praise with nothing specific to build an edit around. Skipping written customer sign-off and publishing based on a verbal go-ahead, which creates real risk if the customer later has concerns about how something was phrased or used.

Building the entire cut around talking-head footage with no product or context b-roll, which reads as flat regardless of how good the content is. And chronological editing that buries the specific result at the end of the video instead of leading with it — most viewers of a 90-second testimonial won't reach the end, so the payoff needs to come first.