Filming a Flop Costs Exactly the Same as Filming a Winner

A lot of business YouTube channels are run on the same strategy. Throw spaghetti at the wall and see what sticks.

A video about whatever felt interesting that week. A topic the founder fancied talking about. Something a competitor posted that looked good.

Nobody with the right skills checking whether the people who might actually buy are searching for any of it. No route from the video to an enquiry. And when nothing sticks, the fix is ‘let’s throw more spaghetti!’

Hope ends up doing the job that strategy should be doing.

But here's the part most business channel owners miss. Filming a high-performing video takes the same effort as filming a flop. Same camera, same setup, same afternoon. If anything the winner costs a little more, because somebody had to work out what it should be about first.

The difference is decided before anyone presses record. It's the research into what that video should be about.

What is a YouTube content strategy for a business?

It's a decision, made before filming, about which questions your buyers are already asking and which of those questions each video will answer. Everything downstream follows from it: the format, the title, the thumbnail, the edit.

Search the phrase and you mostly get production plans. Set goals. Define content pillars. Pick a posting cadence. Batch your filming. All sensible, and all of it tells a team when to film rather than what to film. A cadence is a schedule. It won't tell you whether anybody wanted the video.

Why does publishing more videos stop working?

Because volume only compounds when the videos are aimed at something. Publish weekly for two years with no research layer and you have a hundred videos and no idea which of your buyers' questions you answered, because nobody established what those questions were.

Effort scales. Learning doesn't, unless something is built to capture it.

The cost of getting that wrong shows up clearly on big channels too. Two years of conventional long-form on the channel behind Simon Squibb's YouTube strategy produced 121 videos at a 555-view median. A change of format found roughly fifty times that audience inside five months. The uploads were never the problem.

The quiet version of this fault is vocabulary. Two channels last month were making genuinely good videos, and both described their work in their own industry's language while their buyers searched in ordinary words. Nothing wrong with the content. They were answering questions nobody was typing.

How do you find out what your buyers are searching for on YouTube?

Three places, all free.

• Your own analytics. YouTube's Reach report shows how viewers found your content, and selecting YouTube search shows the terms that brought them there. Most owners I meet have never opened it.
• The search bar, in a private window. Type the start of a phrase your buyer might use and read the autocomplete. Those are real phrasings, not the ones your marketing team would write.
• Your sales calls. The question you've answered thirty times this year is the question your buyers are searching. It's also the video you haven't made.

Stay signed out for that second one. YouTube's documentation says that when you're signed in, the searches you've run and the videos you've watched influence the suggestions you get. You've searched your own topics more than anybody has, so a signed-in autocomplete hands your history back to you dressed as market demand. YouTube also notes that some predictions stay personalised even when signed out, so treat a private window as a cleaner read rather than a perfect one, and discount any suggestion you recognise as your own.

Then two checks before a topic earns a filming slot. Is the demand live now, or did the topic peak three years ago? And are the videos already ranking for it strong and recent, or thin and out of date? A topic can pass the first and fail the second. Skipping that research is what makes a filming day expensive.

What should a business YouTube video send the viewer to?

A page built for the problem that video just raised, not your homepage. One professional services channel had every long-form video pointing at the same homepage link. The content was solid, the views were real, and each video sent the viewers to a page built for everybody and therefore for nobody.

The chain you should be able to complete:

Search → video → a page about that problem → enquiry → client.

Every break in that chain is a viewer who was interested enough to watch and then had nowhere obvious to go. The research decides whether the right person arrives. The link decides whether their interest survives the trip.

How do you know if a YouTube content strategy is working?

Not by views. The test is whether you can name the video that produced your most recent enquiry.

If you can, the channel has a strategy and something measuring it. If enquiries arrive and nobody can say which video sent them, the channel is producing content without learning from it, however good the production has become. Connecting those two things - what you decided to film and what it brought in - is the monthly job behind our groundbreaking YouTube channel management service.

Marisa Peer's channel was run this way. Topics chosen around what her audience was actually searching for, packaged with tested titles and thumbnails, and every video pointing into the funnel rather than at a homepage. It added 70,000 subscribers while I ran it, and the content and pipelines behind it contributed to multiple six-figure revenue. The Marisa Peer case study has the detail.

Do it the other way round. Find the question your buyers are already asking, make the video that answers it, and send that viewer to a page built for that one problem. What it returns to your business is decided before anyone presses record, with the all-important strategy layer.

Then read the result properly. A video that underperforms teaches you something reusable if you wrote down what you expected first, which is why the channel worth comparing yourself to is the one you had last month.