Market gap analysis: a practical guide with real search data

A market gap analysis is a comparison between two things: (1) what customers in a defined market are trying to accomplish and the problems they run into, and (2) what the products currently on offer actually deliver. The delta between the two is the gap.

It differs from a general gap analysis, which is an internal exercise — comparing a company's current state (skills, process, performance) to a desired state. Market gap analysis is outward-facing. The "current state" is the entire competitive set. The "desired state" is defined by real customer language, not by leadership.

It is also narrower than "market research." Market research asks what is happening. Gap analysis asks what is missing — and forces you to name it in a form you could hand to a product or engineering team on Monday.

Before running the exercise on a category you care about, it helps to see what actual search demand looks like for the concept itself. Below is a live comparison of the main terms in the gap-analysis cluster in the US market (source: Semrush, US database, pulled July 2026).

1. 1
   Define the market narrowly

   'Sleep products' is not a market — it's a category. 'Weighted sleep masks for side sleepers' is a market. Narrow enough that two competent researchers would name the same 5–10 competitors independently. If they wouldn't, narrow further.

2. 2
   List 5–10 real competitors

   Include the category leader, 2–3 mid-tier products with real review volume, and 1–2 challengers people talk about but don't buy. Skip anything with under ~50 reviews — the signal is too noisy.

3. 3
   Collect evidence, not opinions

   For each competitor: pull all 1–3 star reviews (the complaints) and the top-voted 4–5 star reviews (what people actually love). Add product page copy, pricing, and — if relevant — one or two forum threads. Aim for ~100+ reviews per product before you trust the themes.

4. 4
   Cluster themes across products

   Tag each piece of evidence with the underlying job or complaint (e.g. 'pressure on eyes', 'strap slips at night', 'washing ruins the fabric'). Then group tags across products. A theme that shows up in 4+ competitors is a market-wide pattern, not a product-specific bug.

5. 5
   Score gaps on 3 axes

   For each recurring theme, score: (a) frequency — how often it appears; (b) severity — one-star language or refund-triggering; (c) coverage — how many competitors already address it well. High frequency + high severity + low coverage = a real gap.

6. 6
   Pick 1–3 gaps and write them as briefs

   For each chosen gap, write: the customer job, the evidence (quotes + counts + sources), why competitors miss it, and what a first version would need to do. If you can't write it in one page, the gap isn't specific enough yet.

Here's what the output looks like on a real category. Numbers below are illustrative of the pattern the method surfaces (based on the sample project shipped inside Prodaro), not universal claims about the sleep-mask market.

Two things to notice. First, the highest-frequency complaint isn't always the gap. "Pressure on eyes" shows up in every product's reviews, but three competitors already address it with contoured designs, so the market gap is smaller than the raw count suggests. Second, a small segment can still be a real gap. Only 3 of 8 products triggered scalp-acne complaints, but zero competitors design against it — a specialist product can own that segment.

Paste the block below into a doc and fill it in per competitor.

MARKET: <narrow definition — who, what job, what constraint>
DATE: <YYYY-MM-DD>

COMPETITOR: <name>
  URL: <product page>
  Price: <$>
  Positioning (from their copy): <one line>
  Reviews analyzed: <count> (source: <Amazon / site / etc>)

  What buyers LOVE (top-voted 4–5★):
    - <quote>  [n mentions]
    - <quote>  [n mentions]

  What buyers HATE (1–3★):
    - <complaint>  [n mentions] [severity: high/med/low]
    - <complaint>  [n mentions] [severity: high/med/low]

  What buyers WISH existed (explicit):
    - <wish>  [n mentions]

---

CROSS-COMPETITOR THEMES
  Theme: <name>
    Appears in: <x / N competitors>
    Coverage: <how many already solve it well>
    Severity signal: <refund language? repeat complaints?>
    Gap score (freq × severity ÷ coverage): <n>

---

CHOSEN GAPS (1–3)
  Gap: <one sentence>
    Customer job: <JTBD phrasing>
    Evidence: <3–5 quotes with source + count>
    Why competitors miss it: <specific reason, not "they're lazy">
    v1 requirements: <3–5 bullets>
    Kill criteria: <what would prove this gap isn't real>

• Mixing evidence with interpretation.

  "Customers want a lighter mask" is interpretation. "37 reviews across 5 products complain the mask is too heavy for travel" is evidence. Keep them in separate columns. If you can't cite a count and a source, it's not evidence yet.

• Confusing product bugs with market gaps.

  One competitor with a broken strap is a QA issue. Five competitors with strap complaints is a market pattern. Never build a product around a gap that only shows up in one competitor's reviews.

• Chasing every gap.

  You will find 15–30 recurring themes on any mature category. Most are not worth building for — the audience is too small, the willingness to pay is too low, or an incumbent will close the gap the moment you launch. Force yourself to pick 1–3.

• Skipping the positive reviews.

  5-star reviews tell you what NOT to break. If every competitor's top-voted review praises 'total blackout,' don't ship a product that compromises on darkness to fix strap slippage.

• Treating the AI summary as the analysis.

  LLMs are good at clustering language and bad at judgement calls. Use them to tag and group. Do the scoring and the gap selection yourself, with the original quotes in front of you.