Launching a CPG Brand at Retail: How to Drive Sell-Through at Kroger, H-E-B and Other Grocers

It’s a Tuesday in March and your broker forwards the first scan report. You’re in 140 stores across the Houston area. The four-pack is doing 0.7 units per store per week, which sounds fine until the broker reminds you the buyer wants roughly double that before the next category review. Thirty-one stores show zero sales in week two. Two of them, you find out later, never set the planogram.

(Illustrative numbers. The shape of the problem is real.)

Everyone told you getting the shelf was the hard part. It’s half of it. Once you’re in, the retailer gives you a few months to prove the item earns its slot, and the only proof it accepts is units moving through the register.

Driving sell-through at Kroger, H-E-B and other grocers comes down to three jobs in the first 26 weeks: make sure the product is actually on the shelf in every store that took it, send shoppers to those specific stores, and show the buyer the lift came from you. Retail media, geo-targeted social, store locator pages, demos and digital coupons are the tools. Velocity is the scoreboard.

We’ve run campaigns for Kroger itself and for Celsius when it needed sell-through at big-box retailers, so this is written from both sides of the shelf.

Velocity is the number your buyer actually reads

Your buyer manages a category, not your brand. Every item in the set gets ranked, and the bottom of the list is where the next new item’s space comes from. You need to know the same numbers the buyer sees, defined the same way.

Two things trip up first-time brands. Adding stores mid-quarter usually pulls U/S/W down, because new stores ramp slowly and the denominator jumps right away. And a handful of zero-scan stores can sink an otherwise healthy average, so the first job is always finding out why those stores are at zero. Usually it’s an unset planogram, product stuck in the back room, or an out-of-stock on the distributor side. None of that shows up in Ads Manager.

Before you spend on ads: slotting, set dates and the promo calendar

Grocery shelf space has always cost money. When the FTC studied slotting allowances back in 2003, it found fees for a national new-product introduction could run from a little under $1 million to more than $2 million, and about one in ten ice cream items didn’t earn enough in their first year to cover their fees, per Supermarket News’ summary of the FTC’s slotting report. Regional deals are smaller and often mix cash, free fill and promotional commitments. Your broker will know the current terms for your category. What matters for marketing is that you’ve already paid to be there, so wasting the first eight weeks hurts twice.

Three things to lock down before any media goes live:

• The set date. Ads that run before the planogram resets send people to a shelf that doesn’t have you yet. Get the reset window in writing and start paid media a week after it, not on it.
• Set verification. Send someone (a merchandising service, a rep, or you in your car) to walk a sample of stores in the first two weeks and photograph the shelf. Spring, The Woodlands, Katy and Pearland stores in one Saturday is doable. If 15% of stores didn’t set, you want to know before you buy geo-targeted ads around them.
• The promo calendar. Temporary price reductions, digital coupons and ad circular features are planned well ahead. Line your heaviest media weeks up with your promo weeks. In accounts we’ve run, a geo-targeted push during a price-reduction week moves noticeably more units than the same spend at full shelf price.

If you’re a Texas brand still trying to get in, H-E-B’s Quest for Texas Best is a real path. The 2026 edition, the program’s 13th, drew a record 978 entries from more than 500 companies, and the grand prize winner got $50,000 plus placement on H-E-B shelves, according to The Shelby Report.

Retail media networks: Kroger Precision Marketing, H-E-B Retail Media and Walmart Connect

Retail media is advertising sold by the retailer, using the retailer’s shopper data, and measured against the retailer’s own sales. Buyers notice who spends in their network. That’s not supposed to affect category decisions, and nobody says it out loud, but it doesn’t hurt.

Where we’d start with a small budget: sponsored product ads on your own items for the category searches shoppers type into the retailer’s app (“sparkling water,” “protein bar,” “electrolyte powder”). Pickup and delivery shoppers search; they don’t browse aisles. If a competitor’s sponsored listing sits above yours on “energy drink” in the H-E-B app, you’re losing trial you already paid slotting for.

Two cautions. First, a retail media report sees every purchase in the loyalty data, so it can claim credit for shoppers who would have bought anyway. Kroger has reported that 75% of targeted onsite display impressions and clicks in a study of thousands of campaigns came from new-to-brand households. That’s Kroger’s own measurement, and it’s still worth asking your rep for new-to-brand and incremental numbers on your campaign specifically. Second, retail media dollars often come out of the same trade budget as your promotions. Decide the split on purpose.

Geo-targeted social around the stores that carry you

Retail media reaches people already shopping the retailer. Social reaches the people who haven’t heard of you yet and live two miles from a store that has you. That second group is where new households come from.

How we set it up:

1. Get the authorized store list. Store numbers and addresses for every location that took the item, from your broker or distributor. Then remove the stores you couldn’t verify.
2. Build location targeting. In Meta Ads Manager, go to the ad set, then Audience, then Locations, and drop pins on each address (or add them in bulk). Meta lets you set a radius anywhere from 1 to 50 miles. In suburban Houston, two to three miles around a store in Cypress or Kingwood is a reasonable start. Inside the Loop, one mile.
3. Split ad sets by retailer. Kroger stores in one ad set, H-E-B in another, so the creative can name the right store and you can read results by banner.
4. Write the creative to send them somewhere specific. “Now at H-E-B. Cold drinks aisle, next to the sparkling water.” Name the store, show the package on the shelf, and give one reason to try it. Vague awareness creative wastes the precision you just paid for.
5. Mirror it on Google. Search campaigns on “[brand] near me,” “where to buy [brand]” and “[brand] H-E-B,” with the same radius targeting, sending clicks to your store locator.

That’s close to what we ran for Celsius: geo-fenced paid social and search around big-box stores in priority DMAs, “find it near you” ads pointing at live store-locator pages, and weekly sell-through data fed back into spend by region. Across 2,850 geo-targeted stores, sell-through in target regions rose 41%, store-locator visits rose 3.4x and cost per store visit fell 37%. The full write-up is in the Celsius case study, and our geo-fencing guide covers the targeting mechanics.

Two rules to check first. Ask your buyer before you put the retailer’s logo in an ad; many chains have brand guidelines and some would rather you run that kind of work through their own network. And if you sell beer, wine or spirits in Texas, read TABC’s marketing practices FAQ before running anything that names a store. TABC treats supplier advertising that promotes a specific retailer as an illegal inducement, while still letting upper-tier members advertise where their products are sold. That line matters for “find it at” ads, so get it reviewed.

Store locator pages that send people to a shelf

Every geo ad and every “where to buy” search ends at your store locator, and most CPG store locators are bad. They show every store a distributor ever shipped to, including ones that dropped you last year, and they don’t say which flavors are where.

A store locator that earns its keep:

• Uses current store-level data. Retailer point-of-sale data through your data provider is best, distributor data is second, and a hand-kept list updated weekly is better than nothing.
• Shows the three nearest stores with distance, which SKUs each carries, and where in the store to look.
• Links to the product page on the retailer’s pickup and delivery site, so “add to cart” is one tap from your ad.
• Shows any live digital coupon with a link to clip it.
• Has retailer pages (/where-to-buy/heb/, /where-to-buy/kroger/) that can rank for “[brand] at H-E-B” searches.
• Fires tracked events on “get directions,” “shop online” and “clip coupon,” which become your conversion events in Meta and Google Ads.

Those three events are the closest thing a CPG brand has to a lead. Tracking them properly lets the ad platforms optimize toward shoppers who looked for a store, not people who watched three seconds of a video. If your site needs the work, that’s part of our web and SEO service.

Online pickup orders count too. The Kroger campaign we ran used recipe creators to drive straight into Kroger pickup and delivery, with every ingredient a tap away. Shoppers who can buy the whole recipe in one cart don’t have to remember your brand name in aisle nine.

Demos and sampling still move product

Some people won’t buy a $3.49 drink they haven’t tasted. A Saturday demo still converts better than almost anything you can buy online, and the effect shows up in that store’s scan data the same week.

Most grocery demos run through demo companies the retailer approves, so your broker or the retailer’s supplier portal is where you start. A few things we’d do differently from most brands:

• Pick demo stores by upside, not convenience. Big-ACV stores with decent foot traffic and weak current velocity. The store closest to your office is rarely the right one.
• Run geo ads around the demo store for three days. “Try it free Saturday at Kroger in Spring, 11 to 3.” A two-mile radius, a small budget, real foot traffic.
• Plan for Houston weather. Cold samples in July sell beverages. A sampling table outside the entrance in August doesn’t. Keep demos inside, near the cooler.
• Measure every demo. Compare the demo store’s next four weeks with a similar store that didn’t get one. If demos don’t show lift, stop paying for them.

Event sampling (a 5K in The Woodlands, a youth tournament in Katy) builds awareness. It only drives sell-through when the nearest stores have product and your locator tells people where they are.

Digital coupons and the promo stack

Kroger and H-E-B both run digital coupons that shoppers clip in the app and redeem with their loyalty account. For a launch, a digital coupon does three things at once: it lowers the trial barrier, it puts your item in front of shoppers browsing coupons, and every redemption lands in the retailer’s data with a household attached.

Brands usually fund these through the retailer’s media network or a coupon partner, and your rep can tell you the current formats and minimums. A few rules from experience:

• Pair the coupon with your geo ads. “Clip $1 off in the Kroger app” gives the ad a reason to act today.
• Use modest values for trial. Deep discounts teach people to wait for the next deal, and your buyer reads base velocity separately from promo velocity.
• Don’t stack a coupon, a price reduction and a demo in the same week unless you want a spike you can’t explain. Space them out so you can tell which one worked.

Measuring lift so the buyer believes it

Your buyer has seen plenty of decks claiming a campaign “drove awareness.” What gets attention is a clean test: stores that got the media against similar stores that didn’t.

The setup we use:

1. Split the authorized stores into test and control groups matched on pre-launch traffic, store size and neighborhood. For a Houston rollout, that might be Spring, Tomball and Atascocita stores as the test group, with comparable stores in Pearland and League City held out.
2. Run the geo-targeted media only around test stores for six to eight weeks.
3. Calculate lift as (test post ÷ test pre) ÷ (control post ÷ control pre) − 1, on U/S/W.
4. Cross-check with the retailer network’s own sales-lift reporting and with Meta’s Experiments tool if you’re running a conversion lift study.

This is how we approached the Kroger work. It ran a recipe-creator program, digital billboards near stores in 38 markets and CTV and streaming radio bought by DMA, then measured store visits. The result was a 24% store-visit lift on 212 million impressions, with a 97% completion rate on CTV. The Kroger case study has the details. A brand can run the same logic at a smaller scale: one city, a few dozen stores, two months.

Put the result on one page for the buyer: velocity in test vs. control, in-stock rate, base vs. promo, new households if the network gives you that, and what you’ll do next quarter.

A 26-week launch plan, with budget pacing

Here’s how we’d pace a regional grocery launch from eight weeks before the set date through the first review.

The percentages are illustrative. The point is to front-load: the trial window decides whether you get a second year, and money saved for month six doesn’t help if the item’s already on the discontinue list. If you’re adding streaming or billboards, our posts on streaming TV ads and Houston billboard costs cover pricing.

Houston is a good place to run this. Kroger’s combined Texas division runs 207 supermarkets across north, east and southeast Texas and southwest Louisiana, and is planning its first new Houston-area store in a decade in Hockley. H-E-B has deep coverage across the region too. Two big grocers in one metro gives you a natural comparison between banners.

Mistakes we see on CPG launches

• Spending on national awareness for a regional shelf. If you’re in 140 Houston stores, a national TikTok campaign mostly reaches people who can’t find you.
• Running ads before the set. The planogram date slips more often than anyone admits.
• Ignoring zero-scan stores. Fix distribution first. Ads can’t sell product sitting in a back room.
• Judging social by clicks to the homepage. Optimize for locator actions and retailer cart clicks.
• Treating retail media as the whole plan. It harvests demand well and creates less of it.
• No control group. Without one, every result is an argument.

Our beverage and eCommerce and CPG pages show the kinds of brands we work with, and if you’re planning creator content for the launch, read our companion piece on creator marketing for beverage and supplement brands before anyone posts. If you’ve got a set date coming and want a launch plan built around your store list, get in touch.

Frequently asked questions

What is a good velocity for a new grocery item?

It depends entirely on the category and retailer. A good velocity is one that ranks your item in the top half of its set, since buyers cut from the bottom. Ask your broker for the category average units per store per week at that retailer, and compare yourself on base weeks, not promo weeks. Also check sales per point of distribution so store additions don’t distort the picture.

How do I advertise my product at Kroger?

Kroger sells advertising through Kroger Precision Marketing, its retail media network powered by 84.51°. Options include product listing ads and onsite display through a self-service platform, plus managed offsite CTV and audio added in 2025. Outside Kroger’s network, brands can run geo-targeted Meta and Google ads around the stores that carry them, sending shoppers to a store locator or the Kroger pickup page.

Does H-E-B have a retail media network?

Yes. H-E-B Retail Media sells promoted-product campaigns on heb.com and the My H-E-B app, plus onsite, offsite, in-store and email placements. It added a self-service option in June 2025, and H-E-B says the network reaches more than 8 million households weekly. Brands can also fund digital coupons and work with the retailer on demos. Ask your broker or H-E-B contact about current formats and minimums.

How long does a new product get on the shelf before it is cut?

There is no universal rule, and timing varies by retailer and category. Most new items get reviewed at the next category review, often six months to a year after the set, and buyers look at velocity, in-stock rate and promo dependence. Plan your heaviest marketing for the first eight to twelve weeks, when trial decides whether repeat purchase has a chance to show up in the data.

How do I measure whether my ads increased in-store sales?

Use a test and control design. Split your authorized stores into matched groups, run geo-targeted media only around the test stores for six to eight weeks, then compare units per store per week before and after in both groups. Lift equals the test change divided by the control change, minus one. Cross-check against the retailer media network’s sales-lift report and any platform lift study you run.

Written by Sean Melton

Sean Melton is the founder of First Byte, a digital marketing agency in The Woodlands, TX. A former global CMO of a $250M software company, he has spent his career inside the businesses he markets. His team runs paid social, search, CTV and creative for brands like Equinox, Kroger and Viceroy, and shares what works here.