Keyword Gaps: A Step-by-Step Method for Finding Missed Traffic on Wildberries and Ozon

A "keyword gap" is a search query where competitors get meaningful traffic while your card does not show up at all or sits below position 50. Finding and closing these keywords is the fastest way to grow on a marketplace. You are not "raising a card from zero" — you are picking up missed traffic for which competitors have already confirmed the demand.

This Support material is a step-by-step methodology for finding keyword gaps on Wildberries and Ozon. The method is used in the standard Trusty MPO audit as the first diagnostic step of the SEO layer. Here it is broken down so that a seller can run the whole procedure alone in 60–90 minutes per card.

The wider context of the methodology sits in the parent materials:

• Pillar "Product card SEO and audits" — the general semantics methodology and the 6 layers of an MPO audit.
• Hub "Wildberries SEO" — WB specifics (Subject ID, the 5 card fields).
• Hub "Ozon SEO" — Ozon specifics (Rich content, category depth).

Sellers often confuse two different tasks:

Collecting new keywords — expanding your semantics to queries that are relevant to the category in principle. Useful, but it still needs proof that the demand is real.

Finding keyword gaps — targeted work with queries whose demand competitors have already confirmed. They get real impressions, clicks and sales on those keywords. Which means:

• Demand does not have to be "created" — it already exists, competitors are serving it.
• The competition is visible — you can see which players take the traffic and how strong they are.
• Queries can be ranked by potential — volume × conversion × your economics.
• The expected effect is calculable — every closed keyword gap delivers a predictable gain.

On a typical card Trusty finds 10–30 keyword gaps; closing them lifts visibility by 20–40% within 4–6 weeks. This is the fastest route to growth, because it requires neither warming up a new cluster nor testing hypotheses.

Keyword gaps appear for four reasons — and each one has its own search pattern:

Reason 1. The keyword is simply missing from the card. The most common case. A competitor mentioned it, your SEO specialist did not. The card is not indexed for it, so it never appears — 0% of the potential traffic on that query. Fixed by adding the keyword to the title, the description or Rich content (Ozon).

Reason 2. The card sits in the wrong category. WB or Ozon placed your card in a subcategory other than the one where competitors rank for your priority keyword. On WB this comes down to the Subject ID, on Ozon to the depth of the category you chose. The card may be perfectly relevant by text, but the algorithm still will not show it for that query because of categorization.

Reason 3. An unfilled filter. A shopper picks "paraben-free cream" from a filter in the search results. If your "composition" attribute is empty, the card disappears. This is not a keyword in the strict sense, but functionally it is the same gap: shoppers who use that criterion never see you. It matters most on Ozon, where a category can carry 30–60 attributes.

Reason 4. Missing synonyms and slang. A shopper searches for "trimmer" while your description says "hair clipper". The WB algorithm will not connect the two on its own (Ozon connects them better, but not always). Slang, colloquial phrasing and regional name variants are frequent sources of missed traffic.

The method works on both WB and Ozon. Platform-specific nuances are flagged separately.

Step 1. Build a list of top competitors

Take 3–5 of your closest competitors in the same subcategory. Not "the biggest names in the niche", but the ones positioned next to you: comparable price, similar product type, the same target audience. Ideally these are competitors you already watch regularly.

In practice: write down 3–5 competitor SKU codes — you will pull their semantics next.

Step 2. Collect each competitor's semantics

For this you need a WB/Ozon analytics service. The basic options:

• MPStats — the most widely used. Open "Product semantics", enter the SKU, and get 50–300 keywords with impressions and positions.
• Moneyplace — an alternative. The "SKU analysis" section.
• JET — for deeper analytics, more expensive.
• The built-in analytics in WB Seller or Ozon Seller — gives less, but it is free.

Export the semantics of each of the 3–5 competitors to CSV. The result is 5 tables with the columns "keyword", "category impressions over 30 days", "competitor position".

Step 3. Collect the semantics of your own card

Same service, same section, only with your SKU. The result is a table with your keywords, impressions and positions.

Step 4. Find the gap: which keywords competitors have and you do not

This is the core stage. Two ways to do it:

By hand in Excel or Google Sheets. Bring the 5 competitor tables together with yours. Using VLOOKUP or MATCH, mark for every keyword whether you rank for it and at what position. Filter the rows where you have no impressions or a position beyond 50.

Through a service. MPStats has a built-in "Semantics comparison" tool: pick your SKU plus the competitors and get a ready-made list of missed keywords.

The result is a list of 30–100 keywords where at least 2–3 of the 5 competitors get meaningful traffic and your card gets none. That is your raw list of keyword gaps.

Step 5. Filter by economics

Not every keyword gap is worth closing. Some of them are:

• Keywords with no real demand (a competitor picked them up by accident, total impressions are low). Drop everything with fewer than 50 category impressions over 30 days — too weak.
• Keywords from the wrong category. If 70% of the impressions for "cream" go to hand cream while you sell face cream, closing that keyword brings impressions and poor conversion. Drop it.
• Keywords at the wrong price point. If the average order value on a query is 500 ₽ and your product costs 5000 ₽, conversion will be minimal. Drop it.
• Keywords that are expensive to promote. High-frequency queries need 100+ buyouts to break into the top 10. If the product margin does not cover that volume, park them for later.

After filtering you are usually left with 10–30 keywords that make economic sense to close first.

Step 6. Rank by potential

Every remaining keyword gap carries its own revenue upside. The base formula:

Potential = (competitor impressions on the keyword over 30 days) × (category average CTR) × (your buyout conversion) × (your average order value)

An example. The keyword "hyaluronic acid moisturizer 50 ml":

• Impressions of the top 3 competitors = 18,000 over 30 days
• Category CTR = 8%
• Your buyout conversion = 12%
• Your average order value = 850 ₽

Potential = 18,000 × 0.08 × 0.12 × 850 ₽ = 146,880 ₽ per month of additional revenue, assuming you close the keyword and land in the top 10 next to the competitors.

The real effect will be lower — usually 30–60% of the calculated potential, because not every card that reaches the results page collects the maximum impressions. But even 50% means 70,000 ₽ per month and up in extra revenue from a single keyword.

Sort the list by potential. The most valuable keyword gaps go on top.

Step 7. Group into clusters

Often 5–10 related keywords close with a single action. For example, "face cream 50 ml", "face cream 50ml", "50 ml face cream" and "moisturizing cream 50 ml" all belong to one cluster, "cream 50 ml", and putting the phrase "face cream 50 ml" into the card title closes all four at once.

Grouping is done by hand (Excel) or through the same MPStats, which has automatic clustering.

The result is a growth map: 5–15 clusters with priorities, total potential and a clear tactic for closing each one.

The specific tactic depends on the cause (see "Where keyword gaps come from").

If the keyword is missing from the card. Add it to the title (if it fits WB's 60 characters), to the description, or to Rich content (Ozon). After reindexing (2 days on WB, 5–7 on Ozon) the card starts appearing for that keyword. To lock the position in you then need the behavioral layer and keyword-targeted buyouts. Details are in the buyouts Pillar, section "Keyword targeting".

If the card sits in the wrong category. On WB, adjust the title and the attributes so the algorithm re-evaluates the Subject ID. In hard cases, raise a support ticket. Details are in the Wildberries SEO hub guide, section "Subject ID". On Ozon, reassign the category in the seller account to the leaf subcategory the top competitors use. Details are in the Ozon SEO hub guide, section "Category and subcategory".

If a filter is unfilled. Fill the attribute in the specifications. On Ozon especially: there are 30–60 attributes per category, and 20–30% of them are often left empty for no real reason.

If a synonym is missing. Add 2–3 key synonyms to the description in natural context. Not as a list in the opening paragraph, but woven into different blocks of the text.

No more than 5–10 clusters per SEO update. The reasons:

• The algorithm recalculates relevance in waves. One big update makes the card float in the rankings for several days before it settles.
• The effect becomes hard to track. If you closed 20 keywords in a single update, you cannot tell what worked and what did not.
• Moderation risk. Abrupt, large description changes sometimes trigger re-moderation.

The baseline rhythm: 5–10 clusters per iteration, 2 weeks of observation, then the next iteration. In 2–3 months that closes 15–30 clusters with a measurable effect.

A few cases where working the list further makes little sense:

• A card with 100% visibility on its priority keywords. If you already hold a large share of the results in the niche, there are fewer growth points than it looks, and the method yields 5–10% instead of 30–40%.
• Systemic problems with the product. If conversion is low because of price, rating or quality, new keywords bring impressions without purchases, which drags the card's overall rating down.
• A recent MPO audit already exists. If the audit was done 1–2 months ago and the current strategy is still running, wait for its results instead of changing course.

Trusty forseller is a free AI tool that automates the first 5 steps of the method on Wildberries. Enter an SKU, and the AI pulls the top 5 competitors in the niche, compares semantics, filters by economics and returns a ranked list of keyword gaps as a PDF in 60–90 seconds.

What the PDF contains:

• The list of keyword gaps with category impressions, competitor positions and a potential estimate.
• The card's Trusty Score across the 6 diagnostic layers.
• A closing plan with priorities for 30, 60 and 90 days.

Limits: Wildberries only (Ozon is in progress), 3 reports per user.

For Ozon cards the methodology is identical, but the process is manual — or you order a full MPO audit with an expert.

• Pillar "Product card SEO and audits" — the SEO methodology as a whole, 5 stages, 6 layers of an MPO audit.
• Hub "Wildberries SEO" — technical breakdown of a WB card, Subject ID, self-check list.
• Hub "Ozon SEO" — indexing of Rich content text, category depth, moving keywords into text blocks.
• Pillar "Buyouts" — once the SEO gaps are closed, you need keyword-targeted buyouts to lock the result in.

Build a list of 3–5 competitors in your subcategory. Compare their semantics with yours in MPStats or Moneyplace. On WB, run the free forseller AI audit and compare the automatic result with what you found by hand — that calibrates the method.

If the task is recurring (a full SKU line, several niches at once), the manual method quickly becomes labor-intensive. In the standard Trusty MPO audit this stage is automated and integrated with the other diagnostic layers: behavioral funnel, ads, reviews and warehouses.