DIY VOC Research vs Hiring It Out

Short version: do it yourself when the question is small, you already know where your market talks, and the decision it feeds can be undone. Hire it out when the decision is expensive, the category is unfamiliar, or you need the answer this week. The trade isn't access. The conversations are public and free either way. The trade is your time and your judgment.

That's an awkward thing for me to write, since I sell the second option. But I'd rather you get this right than buy something you didn't need. Plenty of the brands that email me should have spent a Saturday on Reddit instead, and I tell them so.

What "doing it yourself" actually involves

The reason this comparison gets muddled is that most people costing out DIY are imagining the reading, and the reading is the easy part. Here's the whole job for a single question, like "why do people quit this supplement after a month."

First you find where the conversation lives. That's not one subreddit. It's usually three or four, plus a couple of YouTube comment sections and a review page, and finding them means a lot of searching that goes nowhere. Then you read, and copy out anything that isn't noise. Then you tag: you go back through what you collected and mark every recurring theme, with a tally, so you can tell a pattern from a loud individual. Then you write it into something a decision can be made from.

The step founders consistently underestimate is the third one. Reading feels productive, so people do a lot of it and then stop. Tagging feels like admin, so it gets skipped. But tagging is the only thing standing between a document of quotes and an answer you can defend in an argument with your co-founder.

The alternative used to be a five-figure study

It's worth remembering why DIY became the default for DTC brands in the first place. Buying research meant buying it from a market research firm, at prices built for companies with an insights department.

The Farnsworth Group, a market research firm, says you should plan on $25,000 to $65,000 for a custom qualitative or quantitative project in 2026 (Farnsworth Group, updated February 2026). Drive Research prices focus groups at $7,000 to $20,000 or more per group, and a set of 10 to 15 in-depth interviews at $5,000 to $15,000 (Drive Research, November 2025). For a brand doing a few million a year, that isn't a line item. It's a decision in itself.

So the honest framing isn't "cheap DIY versus expensive service." It's that a third option exists now, sitting between a free weekend and a five-figure engagement, and the question is which of the three your decision deserves. If you want the case against the traditional end of that scale, I made it in why surveys fail DTC brands.

Where DIY research quietly goes wrong

When founders do this themselves and it doesn't work, it's almost never because they didn't try hard enough. It's one of three method problems, and they're worth knowing about even if you never hire anyone.

Stopping before the pattern is real

The most common failure is volume. You read twenty threads, three people say the same thing, and it feels like a finding. It might be. It might also be three people. Qualitative researchers have studied where this line sits: Wutich, Beresford and Bernard put theme saturation at roughly nine in-depth interviews and meaning saturation at roughly twenty-four (International Journal of Qualitative Methods, 2024).

Those are interviews though, where one person talks for an hour. A Reddit comment is a sentence, written sideways, in a thread about something else. You need a great deal more of them to reach the same confidence, which is why "I read a bunch of threads" and "I know what this market thinks" are further apart than they feel.

Reading to confirm

The second failure is subtler and it affects everyone, me included. You come to the research with a suspicion. You find comments that match it faster than comments that don't, because that's how attention works. And because you're the only person reading, nobody catches it.

The fix isn't willpower. It's writing down what you expect to find before you start, then tagging every theme you encounter rather than only the interesting ones. If your prediction was right, fine. If your tally disagrees with your prediction, you've learned something worth more than the hours it took.

Quotes instead of counts

The third failure produces a document that looks like research and can't do the job of research. Fifty great quotes in a doc will not settle a pricing argument, because the other person has fifty different quotes. What settles it is "forty-one of the hundred and twelve people who mentioned price compared us to the pharmacy version, not to the premium brand." That's a number, and numbers end meetings.

What you're actually paying for when you hire it out

Given all that, here's the honest account of what a service adds. It isn't access to secret data. Everything I read is public and you could read it too.

What you're buying is scale, method, and a reader who isn't you. Scale because someone doing this full time works through thousands of comments in the time a founder works through two hundred. Method because the tagging is systematic rather than improvised. And distance because I don't have a theory about your brand that I'm quietly hoping to confirm, which is worth more than it sounds.

You're also buying the calendar. That's the part brands undervalue until they're up against a launch date. Fifteen hours doesn't sound like much until you try to find it inside a week where you're also the person doing customer service, the ads, and the supplier calls.

Side by side

Notice the fifth row, because it cuts the other way. You know your category better than any outside researcher will after a week of reading. When the question is narrow and the market is one you've lived in for three years, that advantage can outweigh everything in the right-hand column.

When DIY is genuinely the right call

Four situations, and I'd say the same thing on a sales call.

The question is narrow. "What do people call this problem" is a couple of hours of reading, not a study. The decision is reversible, so a wrong answer costs you a test rather than a quarter. You already know the two or three places your market talks, which removes the most annoying part of the job. Or you've never done this before and want the skill, in which case doing one yourself is the only way to get it. The complete guide to VOC research for DTC brands covers the method end to end, and how to use Reddit for DTC research is the best single platform to start on.

When hiring it out earns its keep

The mirror image. The decision is expensive and hard to reverse, like a launch, a repositioning, or a price change. You're entering a category you don't know, where you can't tell a representative complaint from an outlier. You have a deadline that doesn't move. Or, most often in practice, you've tried it yourself, got a document of quotes, and realised you can't tell what it means.

There's also a volume argument that has quietly got stronger. Reddit reported 130.3 million daily active uniques in Q2 2026, up 18% year over year (Reddit Q2 2026 earnings call). More conversation is good for research and bad for skimming. The bigger the haystack gets, the more the systematic approach pulls ahead of the weekend one.

And the cost of being wrong keeps climbing. Ecommerce customer acquisition cost ran $68 to $84 in 2025, up 40 to 60% from 2023 (Swell, DTC Ecommerce Statistics, 2026). When every customer costs that much to win, pointing good ad spend at a misread market is expensive in a way it wasn't three years ago.

The middle path most brands should take first

Here's the thing I've noticed across the brands I've worked with, and it's the recommendation I'd actually give. Do one yourself first, on a question that doesn't matter much, and time it honestly.

Two things come out of that week. You find out whether the work suits you, and some founders love it and never need to hire anyone again. And you find out what fifteen hours of your attention is worth against everything else it could have gone to. That's not a number I can tell you. It depends on what else is on your desk that week, and it changes as the business grows.

The founders who make this decision badly are the ones who never run the experiment. They either outsource on reflex, and lose the category instinct that would have made them a better reader, or they insist on doing everything in-house for years while the reading quietly never happens. Either way the decision got made by default, which is the one thing this whole discipline is supposed to prevent.

It's also worth saying the two aren't permanent camps. Most of the brands I work with do their own reading between projects and bring me in for the calls that scare them. That's the correct shape. And once you have the answer, from either route, the next problem is turning it into copy, which is using customer language in your DTC marketing.

Frequently asked questions

Should I do VOC research myself or hire someone?

Do it yourself when the question is small, you already know where your market talks, and the decision it feeds is reversible. Hire it out when the decision is expensive, the category is new to you, or you need the answer this week rather than next month. The trade is time and judgment, not access. The conversations are public either way.

How long does DIY voice of customer research take?

For one focused question, budget 10 to 15 hours. Roughly 2 to 3 hours finding the right threads, 4 to 6 hours reading and pulling comments, 3 to 4 hours tagging what repeats, and 1 to 2 hours writing it into something you can actually use. Most founders underestimate the tagging step, which is the one that decides whether the output is research or a pile of quotes.

How many customer comments do you need before the answer is reliable?

More than most people stop at. Wutich, Beresford and Bernard (2024) put theme saturation at around 9 in-depth interviews and meaning saturation at around 24. Public comments are shorter and noisier than interviews, so you need considerably more of them, which is why reading twenty threads and calling it done tends to produce a confident wrong answer.

Is DIY research cheaper than hiring a market research firm?

Against a traditional firm, yes, by a wide margin. The Farnsworth Group puts a custom qualitative or quantitative study at $25,000 to $65,000 in 2026, and Drive Research prices focus groups at $7,000 to $20,000 per group. That gap is what made DIY the default for most DTC brands. The newer question is whether your own time is the cheapest way to close it.

What usually goes wrong when founders do their own customer research?

Three things. They stop reading too early, so a loud minority looks like a pattern. They read looking for the answer they already suspect, and find it. And they collect quotes without tagging them, which produces a document that feels like research but cannot settle an argument. All three are fixable, and all three are about method rather than effort.

What to do next

Take the question that's actually bothering you right now and ask two things about it. If I get this wrong, what does it cost, and can I undo it? If the answer is "not much" and "yes," block out a Saturday and do it yourself. You'll get an answer and you'll get the skill, and the skill is the part that keeps paying.

If the answer is "a lot" and "no," then the only real question left is whether fifteen hours you don't have is the cheapest way to buy certainty. Sometimes it is. Often it isn't. Either way, it beats deciding on a hunch, which is still how most of these calls get made. Qualtrics surveyed more than 700 senior marketing and insights executives in mid-2025 and found two-thirds still rely on gut instinct for critical decisions (Qualtrics, 2025). DIY or hired out, both beat that.

Sources

1. The Farnsworth Group. (2026). How Much Does Primary Market Research Cost in 2026 (updated February 12, 2026). Link Retrieved August 2026.
2. Drive Research. (2025). How Much Does Market Research Cost? (published November 26, 2025). Link Retrieved August 2026.
3. Wutich, A., Beresford, M., & Bernard, H. R. (2024). Sample Sizes for 10 Types of Qualitative Data Analysis: An Integrative Review, Empirical Guidance, and Next Steps. International Journal of Qualitative Methods. Link Retrieved August 2026.
4. Reddit, Inc. (2026). Q2 2026 Earnings Call Transcript (130.3 million daily active uniques, up 18% year over year), July 30, 2026. Link Retrieved August 2026.
5. Qualtrics. (2025). Marketing Execs Drowning in Data Still Rely on Gut Instinct (700+ senior marketing and insights executives, surveyed July to August 2025). Link Retrieved August 2026.
6. Swell. (2026). DTC Ecommerce Statistics (ecommerce CAC of $68 to $84 in 2025, up 40 to 60% from 2023). Link Retrieved August 2026.