The questions store owners ask, answered first.
Each guide starts with the answer, then the numbers and their sources, then what to do about it in your store.
- Answer in the first paragraph
- Every number with its source
- What the agent does about it
Pick your question
Grouped by topic.
Retention
Keeping a customer costs 5 to 25 times less than winning one
The 5x figure is from 1990. The range today is wider, acquisition costs keep rising, and most stacks still treat every visitor as a stranger.
Read the guide →AbandonmentYou have five abandonment problems. A cart flow solves one.
Cart abandonment is the last failure point, not the first. Search, browse, category and wishlist abandonment lose more, and a standard email flow does not see them.
Read the guide →EmailEmail deliverability: how to reach the inbox, not the spam folder
Accepted is not delivered. Gmail, Yahoo and Microsoft now require authentication, one-click unsubscribe and low complaints. Then engagement decides the rest.
Read the guide →EmailGmail's Gemini summaries: who you email matters more than how
Gmail now reads your email before your customer does, and shows how often you send. Format for the summary, then send fewer emails to the people who want them.
Read the guide →AnalyticsFive marketing decisions to take from data, not instinct
Send time, discount type, segments, frequency and recovery timing have no universal right answer. Your own customers' behaviour has one, and a holdout test shows it.
Read the guide →SeasonalValentine's Day campaigns: start early, segment, win the last days
Valentine's Day is three campaigns, not one: an early preview, gift guides for each kind of buyer, and a last-minute push built around shipping deadlines.
Read the guide →Personalization
Ecommerce personalization: what it is and what works
Most personalization shows the nearest similar product. The kind that pays shows each shopper the next thing they will buy, and works toward the goal you choose.
Read the guide →RecommendationsMost recommendation engines retrieve. They do not recommend.
The two-tower model from Google's 2016 YouTube paper finds the nearest similar item. It cannot know what a customer is worth or what they will buy next. That is the layer that pays.
Read the guide →SearchPersonalized site search: the first screen decides the sale
Keyword search shows every shopper the same hundred results. Personalized search puts her size, her brands and her price range on the first screen, then tells the rest of the store what she looked for.
Read the guide →Anonymous visitorsHow to reach the visitors who never log in
Start the profile at the first click, not the first login. Then the site sells to every visitor, and nothing is lost when they finally give an email.
Read the guide →Tracking and attribution
Server-side tracking: see the orders your pixel misses
A browser pixel fails whenever the browser says no. Server-side tracking records the event where it happens and sends it to the ad platform directly. Then the question is what you send.
Read the guide →Decision intelligenceA decision intelligence platform decides what to do for each customer
Gartner published its first Magic Quadrant for the category in January 2026. For a B2C brand it means one goal for the site, the messages and the ads, decided per person, and proven with a holdout.
Read the guide →AttributionThe ads you skip still sell. Your attribution cannot see them.
Exposure builds preference without memory, weeks before the order. A 7-day window credits the last click and starves the channels that created the demand.
Read the guide →Customer dataA customer data platform only pays when the profile acts
Most CDPs collect and report. A store needs one profile per shopper, anonymous visitors included, that picks the products, times the messages and builds the ad audiences.
Read the guide →PrivacyThird-party cookies: Chrome kept them, your store should not rely on them
Google dropped its plan to remove third-party cookies, but Safari and Firefox restrict them and consent rules cut them further. First-party data is what still works.
Read the guide →Strategy
Stop paying ads to find customers you already have
Your ad platform counts a returning customer as a new acquisition. Exclude known customers, predict who is ready to buy, and reach them on channels you own.
Read the guide →Enterprise stackFive failures of the enterprise marketing stack, and the one cause
Email optimizes opens, the site optimizes clicks, ads optimize fast conversions. Nothing in the stack optimizes what each customer will be worth, and the revenue goes missing in between.
Read the guide →Marketing stackFive blind spots in an ecommerce marketing stack
Every tool in the stack optimizes its own metric, so data and revenue fall through the gaps between them. Five tests you can run on your own data this week.
Read the guide →StrategyThe 4 Ps of marketing, revisited: win on a few, not all
Georgi Georgiev, Releva's CEO, on why the marketing mix is not a list to optimize all at once, and which Ps add value in ecommerce.
Read the guide →The numbers, once a month
What the stores on Releva saw last month: Viber against email, email benchmarks, Meta ads, the season. Shares and ratios you can compare with your store, one email a month, no sales pitch.
Thanks. The next numbers reach you at the start of the month.
One email a month with the numbers, and now and then other news from Releva. Unsubscribe any time. Privacy policy
See which of last week's visitors you missed
One call, with your store and ad accounts open. Then 30 days free. A holdout group decides: if the agent doesn't add orders in 30 days, you don't pay.
See it on your store
You'll talk to one of us.