The Outlier Method: How to Pick YouTube Topics for B2B SaaS

The outlier method is a way to choose YouTube topics from data instead of opinion. You compare each video's views with the median of the channel's recent uploads, and treat the videos that clearly beat the median as evidence that people want an answer to that question. It is the cheapest topic research there is, and it is how we plan every channel.

Why the median, not the average

Averages lie on YouTube. One viral video drags the average far above what a normal upload gets, and hides everything else. The median is the middle value when you sort the views: it shows what a typical upload does on that channel. A video that earns several times the median did something different, and that difference is worth studying.

What it looks like in public data

We compared the top video on each of several B2B channels with the median of that channel's recent uploads (checked 27 October 2026). The gap ranges widely: from about 64 times the median on Surfer Academy to about 6 times on Firecrawl. A gap that wide is the signal. These videos are not normal uploads. They are answers to questions that a lot of people had. We show the same comparison in each case study.

The method, step by step

1. Pick five channels. Yours, two direct competitors, and two channels that sell to the same buyer but are not competitors. Adjacent channels are often the best source of ideas because they are not chasing the same keywords.

2. List the last thirty uploads for each. Note the title, views and publish date. Thirty is enough to get a stable median without going back to a different era of the channel.

3. Find the median for each channel. Sort by views and take the middle value.

4. Mark the outliers. Anything at roughly three times the median or more, and anything that is the best by a wide margin in a recent run.

5. Read the titles and watch the first minute. What question does each one answer? What does the viewer get? What is shown on screen?

6. Group the outliers. They usually cluster around two or three themes. Those clusters are your first topic list.

7. Write your version. Same question, your product, your angle, and a narrower buyer if you can.

What to ignore

• Spikes from news events or a one-off launch.

• Giveaways, contests and collaborations that borrowed another audience.

• Videos with paid promotion behind them. You cannot see it, so treat extreme outliers with caution.

• Topics unrelated to a problem your product solves. Views are not the goal. Demos are.

Turning an outlier into a video

A strong outlier proves demand for the question. It does not tell you how to answer it. Write the question as a buyer would type it, decide what you can show that a competitor's video cannot (a real workflow, a real number, a customer), and make the opening state the answer. Then check the result against your own median after two weeks. If it beats it, make the next one in the same cluster.

How often to repeat it

Run the full exercise when you plan a quarter. Then run a lighter version monthly on your own channel. Your own outliers are the most useful because they show what your audience wants from you, not from a competitor.

A worked example from public data

Here is the method applied to three real channels, using public numbers checked on 27 October 2026. It shows how large the gap can be, and what to take from it.

• Surfer Academy has a recent-upload median of about 5,100 views. Its top video, a complete SEO and AI SEO course for 2026, has about 324,000, roughly 64 times the median.

• Rho has a median of about 14,000. Its top video, about how funding works for startups, has about 678,000, roughly 48 times the median.

• Passetto has a median of about 196. Its top video, how to build a go-to-market strategy for B2B SaaS, has about 4,000, roughly 20 times the median.

Three things stand out. All three outliers are broad, practical topics that many people search for. Two are long, complete treatments of a subject. And the outlier appears even on a tiny channel: Passetto has 311 subscribers, yet one video earned about twenty times its normal views. That is the practical point of the method. You are looking for questions with more demand than your channel currently reaches.

A spreadsheet you can build in twenty minutes

You do not need a tool. Make one sheet per channel with these columns:

1. Title

2. Views

3. Publish date

4. Length in minutes

5. Format (how-to, comparison, opinion, story, review)

6. Multiple of the median (views divided by the median)

7. The question it answers, in your words

Sort by the multiple. Read the top five rows of each sheet, then put all the questions from all five sheets in one list. Duplicates and near-duplicates are your strongest candidates, because several channels found the same demand.

Setting the threshold

Three times the median is a sensible starting line for a channel with steady uploads. Raise it to five times if the channel has erratic results, and lower it to two times if the channel is small and uploads are consistent. The threshold is a filter, not a law. What matters is that you read the titles above the line and ask why those videos beat the rest.

From outlier to a brief

For each topic you choose, write a one-page brief before anyone films:

• The buyer question, worded the way a buyer would type it.

• Who it is for, in one line, with their role and situation.

• The outlier evidence, with the channel, the title and the multiple.

• Our angle: what we can show or say that the original does not.

• The proof moment: the screen, number or customer that makes the answer believable.

• The call to action: the one next step we want the viewer to take.

The brief takes fifteen minutes and prevents the most common failure: a well-made video that answers a question nobody asked.

Reading your own channel

When you have ten or more uploads, run the method on yourself. Look for patterns in what beat your median:

• Which formats do best, and are you making enough of them?

• Which topics cluster at the top?

• Do longer or shorter videos win?

• Which ones led to demo requests, regardless of views? Low-view videos on bottom-of-funnel questions can matter more than a large one.

Views and demos do not always move together, so keep both lists and use the second to override the first.

Common mistakes

• Using the average. It hides the pattern you are looking for.

• Copying the format as well as the question. The question is the proof of demand. The format should serve your product.

• Studying only direct competitors. They are often chasing the same topics. Adjacent channels show you gaps.

• Chasing views at the expense of fit. A topic that draws a large audience who will never buy is a vanity win.

• Running it once. Demand shifts. Repeat it every quarter.

How many topics you need

Enough for a quarter. For one video a week that is about twelve, plus a few spares. Rank them by how directly they sit in front of a buying decision, then by the strength of the outlier evidence, then by how easily you can show the answer on screen.

What to do when your niche has few channels

If you sell into a narrow category, you may find only one or two relevant channels. Widen the search in three directions rather than giving up:

• Sideways: channels that serve the same buyer with a different product. A payroll tool and a recruiting tool share buyers and questions.

• Upstream: channels about the job the buyer does, not the software they use. The finance lead who buys your tool watches videos about finance.

• Search itself: type the questions a buyer would ask into YouTube and note which videos come up first and how many views they have. Strong results with low competition mean the question is open.

The numbers will be noisier on small channels, so lower your threshold and rely more on the titles than on the multiples.

Turning a quarter of topics into a calendar

With twelve topics ranked, assign them to weeks so that the schedule has variety and a purpose:

1. Start with two or three bottom-of-funnel videos. They are the ones your sales team can use immediately.

2. Alternate formats so the channel does not feel repetitive: a how-to, then a comparison, then a point of view.

3. Place the strongest outlier topic in week three or four, once the team has found its rhythm.

4. Hold two slots open for topics you spot while the quarter runs. Your own early results will suggest them.

Review the plan at week six. Drop anything that no longer fits and replace it with the best of your new candidates.

Next steps

Once you know what to make, the question is how to get it seen. Read the 14-day distribution playbook, or see what 60 top B2B titles have in common to write titles that earn the click.

This is the method inside the B2B YouTube Engine. Book a call and we will run the outlier check on your category before you commit to anything.

Data: public YouTube statistics, checked 27 October 2026. Medians are from each channel's recent uploads.