A founder at a desk facing a wall of sticky notes grouped into a few clusters, with a phone showing a blurred social feed and an open laptop

The short answer: Most consumer brands pick a channel first and look for buyers second, and they pay for it in wasted spend. With Meta’s average price per ad up 14% from a year earlier, every impression in front of the wrong person costs more. Write your core buyer down on one page, pick the two channels where that buyer already pays attention, and run small tests with a stopping rule before you spend real money.

Ask a founder who their customer is and you often get a channel instead. “We’re a TikTok brand.” “We’re big on Amazon.” “Our customer is on Instagram.” Those are places, not people. When the channel comes first, the brand ends up paying the platform to figure out who the buyer is, one expensive impression at a time.

This post is about doing it in the right order: define the buyer for the whole business, then choose the few channels that actually reach that person. It goes wider than our post on reaching the right shoppers for a new retail launch, which covers one set of stores. This is the audience behind every launch.

A lesson from sorority rush

Look at what brands did around college rush season this year. Front Office Sports reported on Aug. 28 that Nike sponsored rush for at least six sororities, at schools including South Carolina, Texas and Arizona State, with members receiving product, and that Alo Yoga and Set Active have also gone after rush. Net Influencer reported on Sept. 14 that CampusLaunch, a company that sets up these campus programs, ran 32 activations in 30 days, each with 150 to 200 chapter members posting, and that one BODYARMOR event at Ole Miss drew about 300,000 views on its main video.

The New York Times showed the student’s side on Nov. 15: a University of Arizona sorority junior with more than 45,000 TikTok followers, a talent manager and several paid brand deals, the biggest worth $3,000, plus a mailroom full of free products from brands. Sisters at several sororities there serve as brand ambassadors, posting about free products, hosting events like yoga on the lawn with Alo, and modeling brand merchandise with discount codes. By brand standards, those are small checks.

Whatever you think of the tactic, notice the order. These brands did not start with “we should be on TikTok” and hope the right people saw it. They started with a narrow buyer, college women, mostly 18 to 22, and went to the place and the moment where that buyer gathers and listens to each other. The channel was the answer to a question about people.

Most brands do not need a sorority. They need the same discipline: know exactly who buys, then go where that person already is.

Why this matters more right now

Reach is getting more expensive. In its third quarter 2026 results (Oct. 21), Meta reported that ad impressions rose 10% from a year earlier and the average price per ad rose 14%. In the second quarter the price per ad was up 12%. More demand for the same attention means a higher price for every view.

The platforms are also taking more of the targeting into their own hands. Marketing Brew reported on Oct. 8 that media buyers are working around higher costs with Meta’s automated Advantage+ campaigns and constant creative testing, with one buyer summing up Meta’s message as “creative is the new targeting.”

Here is what that means for a small brand. If the algorithm picks the audience from your creative, then your creative has to be written for one clear buyer. A vague ad gets shown to a vague crowd, and you pay more for it every quarter. A clear buyer is now the cheapest targeting tool you have.

Step one: define the buyer on one page

Skip the persona with a stock photo and a made-up name. A useful buyer sheet answers questions that change what you do:

That last trust question matters more than most brands think. A youth survey from retailer Pacsun and GlobalData (reported Sept. 14; a retailer’s own study of 6,126 people ages 11 to 24) found TikTok was the top place they discover brands, but for style advice they trusted friends (34%) and family (27%) far more than influencers (10.8%). Discovery and trust can live in different places. The sorority programs work because they put the product in the hands of peers.

Fill the sheet from evidence, not from memory. Your reviews, customer service emails, DTC order data and retailer shopper reports already hold most of the answers.

A woman in her 30s in a kitchen checking her phone while holding an unbranded purple supplement bottle, a grocery bag of vegetables on the counter beside her

Step two: pick two channels, not seven

Small brands lose money by being a little bit everywhere. Score each possible channel against your buyer sheet with four questions:

  1. Does my buyer actually spend time there, by the evidence on the sheet, not by the hype?
  2. Can I test it for an amount I am willing to lose?
  3. Can I measure it well enough to tell a win from noise in a few weeks?
  4. Can I keep it up with the people, product and cash I have for a full year?

Run the usual list through it: in-store sampling and demos, retailer and Amazon search, paid social, creators and micro-influencers, email and text, search, and community groups and events. Keep the two that score best. Everything else waits until those two are working.

One channel most brands under-rate: being findable when the buyer goes looking. If a creator or a friend sends someone your way and your product page or listing cannot answer their question, you paid for a visit and lost the sale. Our post on how CPG brands get found online has a one-hour check for that.

Our founder directed brand, paid search, SEO and SEM for Vitamin Advisor while its ecommerce sales grew from $1M to $10M. Our view from that work: paid media pays back for the brand that already knows exactly who it is talking to, and punishes the one still trying to find out.

The audience plan you can run this month

Cheapest steps first. Most of this uses data you already own.

  1. Pull who buys today (one afternoon, free). Export 12 months of DTC orders, read the last 100 reviews, and pull any retailer shopper insights you have. Note age bands, household, location and the words customers use.
  2. Write the one-page buyer sheet (one afternoon, free). Use the seven questions above. If two very different buyers show up, write two sheets and pick the bigger, more profitable one to lead with.
  3. Ask ten customers how they found you (one week, free). A short call or email with one question: “Where did you first hear about us, and what made you buy?” Ten honest answers beat a stack of trend reports.
  4. Score the channels and pick two (two hours, free). Use the four questions. Write down why you chose each one so you can check the reasoning later.
  5. Seed peers before you pay for reach (two to four weeks, low cost). Find your version of the sorority chapter: a run club, a gym, a parents’ group, a hobby community. Give product to people your buyer already trusts, and require a clear disclosure, as our micro-influencer post explains.
  6. Run one small paid test per channel with a stopping rule (two to four weeks, moderate cost). Set the budget, the success number and the end date before you start. An illustration, not a real brand: $1,500 per channel, one message written for the buyer sheet, and a rule to stop if cost per new customer is more than you earn on that customer’s first two orders.
  7. Time the bigger spend (next quarter). Marketing Brew’s buyers note that peak ad prices “do come down” after the holiday rush. If your buyer does not need you for gifting, January can be a cheaper time to scale what worked.

What not to do

Next step

Reaching the right audience is the front half of being found. The back half is making sure that when your buyer goes looking, search, retailer sites and AI assistants can actually answer for you. Our pillar post, Why Your CPG Brand Isn’t Being Found: The New Digital Discovery Stack, shows how the pieces fit. If you want a quick read on where your plan has gaps, the free Launch Scorecard below takes about three minutes.