The video looks good. People watched it. Someone has already suggested making a series.
Before approving episode two, I would ask what episode one was supposed to change. Did it help a buyer understand the product? Bring qualified visitors to an offer? Reduce repetitive sales explanations? Help a customer use something they had already bought?
“We need more video” skips that question. Video is a format. Its business value depends on the job you give it.
Write the job before the script
A product demonstration and a brand introduction need different evidence. The demonstration might aim to improve the proportion of qualified prospects who request a trial. A brand introduction might aim to generate additional demand among people unfamiliar with the business. A training video might aim to reduce support work without hurting customer outcomes.
Choose one primary job, the audience, and the action or change you expect. Put a review date on it. If nobody can agree on those details, the production brief needs more work.
Be careful with videos that promise to do everything. A single asset can serve several purposes, but that does not entitle it to claim every sale, every saved hour, and every new subscriber it touches.
Views tell you something. Find out what.
YouTube Analytics shows how thumbnail impressions lead to views and watch time. Those are useful clues about distribution and attention. They do not tell you whether additional customers bought at a worthwhile cost.
A large view count could mean the right audience found a useful answer. It could also mean the opening attracted people who will never need your product. An average view duration needs context: what did viewers need to see, and did they reach that point?
YouTube’s audience retention report shows how different moments held viewers’ attention. If people leave before the demonstration starts, try moving it earlier. If they watch the answer and leave before a long closing sequence, perhaps they got what they needed. A drop is a question to investigate, not automatically a creative failure.
Connect attention to the intended job. Track the next step where practical, such as a relevant landing-page visit, a trial request, or a sales conversation. Ask whether those actions came from suitable buyers and whether they progressed.
Count the cost of getting it used
The production invoice is only the beginning. Include planning, writing, preparation, employee appearances, revisions, captions, distribution, paid promotion, tools, and measurement. Staff time is a cost even when nobody sends an invoice for it.
A video that sits on a page nobody visits has a distribution problem. Another shoot will not solve that. Decide where the asset will appear and how the intended audience will encounter it before paying to produce it.
For a reusable video, choose a reasonable evaluation period. Report results to date separately from anticipated future benefits. Do not make today’s economics look better by assuming years of useful life that have not happened yet.
Ask what changed because the video existed
Suppose, hypothetically, a demonstration costs $8,000 to produce, $2,000 in staff time, and $2,000 to distribute. Full cost is $12,000.
A well-designed comparison of otherwise similar prospect groups estimates 30 additional purchases during the review period. Each purchase provides $600 in contribution after variable fulfillment costs. Incremental contribution is $18,000. ROI is ($18,000 minus $12,000) divided by $12,000, or 50%.
If the plausible range is 10 to 40 additional purchases, the implied ROI ranges from negative 50% to positive 100%. That is useful uncertainty. It tells you how much the conclusion depends on the estimated lift.
A randomized comparison, when practical and large enough, provides stronger evidence than simply comparing people who chose to watch with people who did not. Volunteers may already be more interested. Google makes the same underlying distinction in its Conversion Lift documentation, separating additional conversions from ordinary attributed results. Access to a particular tool is separate from the principle.
Decide what deserves another dollar
Keep the video if the business evidence supports it. Improve the opening if relevant viewers abandon it before the useful part. Fix distribution if qualified people never encounter it. Stop commissioning similar work if the cost remains too high for the likely benefit.
For a support video, calculate actual avoidable workload or capacity value carefully. Ten saved hours do not become ten hours of cash savings unless costs actually fall. They may still matter if employees can use the time for other valuable work. Report that benefit separately and avoid counting it again in customer contribution.
The decision is not whether video works in general. It is whether this video, doing this job, produced enough additional value to justify its full cost, and what the evidence suggests you should do next.
Use the marketing ROI calculation guide to work through the economics and the guide to attribution and incremental results to examine the evidence. Explore the free marketing resources, or work with me on your strategy and implementation.
For the next decision, read Marketing ROI vs. ROAS: Your Ads Can Win While You Lose Money; Where Should a Small Business Spend Its Marketing Budget?.
