Google Ads guide

How to Set Up Google Ads Keyword Match Types for Startups

To set up Google Ads keyword match types for a startup, document the customer and query job, map each theme to a page and product path, choose broad, phrase, or exact for a specific discovery or governance role, add negatives from real fit boundaries, and connect search terms to qualified activation, pipeline, or another approved outcome. Make the decision record before changing the live account.

Google’s keyword matching guidance describes current broad, phrase, and exact behavior, including meaning-based matching and overlap. Check the live account controls and current documentation before applying the setup.

Step 1: Write the customer and query map

Create a short row for each query theme. Include the target company or user, problem, product, market, funnel stage, page, next action, owner, and business outcome. Start with themes such as problem discovery, category, product or feature, competitor comparison, brand, support, jobs, and education, then remove themes the startup cannot serve today.

Do not let a keyword list define the customer by accident. “Project management software” may need category proof, while “how to manage projects” may be education or problem discovery. A branded product query may need a fast signup path, while a security or procurement query may need documentation and a sales owner.

Choose the event that proves value for this setup. It might be an activated account, connected integration, qualified signup, accepted lead, meeting held, opportunity, paid conversion, or retained account. Keep clicks and basic forms as diagnostic events unless they are the outcome the startup has explicitly approved.

Step 2: Map themes to pages and measurement

Give every theme a page that answers the query and supports the next action. Confirm that the ad, page, product version, pricing, implementation promise, security information, signup flow, and sales handoff agree. A startup should not use broad discovery to send unrelated problem questions to a generic homepage.

Before adding keywords, verify the measurement path:

  • product-led flows should distinguish account creation, activation, meaningful usage, retention, and paid conversion;
  • sales-led flows should distinguish form, accepted lead, meeting, opportunity, stage advance, and contract;
  • hybrid flows should preserve the join between acquisition, product activity, CRM status, and revenue; and
  • all flows should mark spam, duplicates, unsupported accounts, delays, and unmatched records.

Use Google’s conversion measurement guidance for the platform layer, then reconcile it with product analytics or CRM data. Do not optimize the match-type decision against a conversion event that the startup cannot validate.

Step 3: Choose the initial match-type role

Use broad for a problem or category discovery theme when the page is specific, customer boundaries are explicit, measurement is reliable, and a person can review search terms. Broad is a way to learn market language, not a replacement for a target customer or a landing page.

Use phrase or exact for a proven product, category, brand, competitor, or high-value use-case theme when the team wants a clear review unit. Current behavior is meaning- and context-based, so do not describe either as literal-only protection. Pick the role based on the decision the team needs to make, and record why.

For every keyword or theme, save the campaign, match type, page, customer boundary, market, budget, conversion action, negative scope, and promotion rule. A startup with sparse data needs this context because a few clicks or leads should not be mistaken for a stable pattern.

Step 4: Add negatives from real fit boundaries

Review the startup’s first search-term set and write exclusions that a teammate can explain. Possible boundaries include jobs, courses, free-only requests, support, unrelated meanings, unsupported industries, out-of-market locations, account sizes outside the service model, and product features that are not available.

Do not exclude a term only because it is unfamiliar or has low volume. Check its meaning, customer, page, product fit, and expected outcome. “Open source,” “API,” “enterprise,” and “template” can describe valuable demand for one startup and irrelevant traffic for another. Keep the negative reason and apply it at the narrowest scope that matches the evidence.

Test the effect of shared negative lists. A rule that is correct for a sales-led campaign may remove useful self-serve traffic, and a support exclusion may block a product education path that the startup intentionally supports. Review negative changes again when the product, pricing, plan limits, or target market changes.

Step 5: Organize campaigns around decisions

Keep themes together when they share the same page, customer boundary, budget logic, conversion event, and next action. Split them when a change would need a different page, owner, offer, sales path, product version, or economic judgment. This is more useful than creating a separate container for every spelling variation.

For example, broad problem discovery can have a page and review rule focused on market language, while a proven category theme can have a different page and qualification path. Brand navigation may be efficient but should not silently be used as proof that non-brand discovery is working. Competitor research can need a distinct policy, message, and risk review.

Save the setup before launch and whenever it changes:

  1. query theme and customer job;
  2. keyword and match-type role;
  3. page, product capability, and next action;
  4. market, customer boundary, and campaign owner;
  5. conversion action and later-stage quality event;
  6. negative list and scope;
  7. budget, promotion, and pause rule; and
  8. change reason and review date.

Step 6: Review search terms and make one decision

Inspect terms by problem, category, feature, product, brand, competitor, role, industry, location, support, jobs, price, and customer stage. For each useful pattern, choose one response: promote it to a proven theme, add a negative, move it to another campaign, improve the page, correct the product or qualification path, or leave it under observation.

Review the account’s other changes at the same time. Product launches, pricing, free-plan limits, sales response time, page revisions, tracking fixes, seasonality, and budget changes can all alter the result. Keep the search-term review from claiming a match-type effect that the available evidence cannot isolate.

Step 7: Check activation and pipeline quality

Join the query theme to clicks, signups, activation, qualified account, accepted lead, meeting, opportunity, paid conversion, retention, expansion, and cost. Use the stages the startup actually has, and show the lag between acquisition and later quality. A keyword that produces fewer leads but more activated or retained accounts may be more useful than one that produces cheap forms.

Compare similar markets, product versions, offers, customer boundaries, and maturity windows. Note sample size as evidence, not as an invented pass/fail threshold. If the startup does not yet have enough later-stage data, report that uncertainty and use the available signal with a planned quality review.

Step 8: Decide whether to expand, change, or pause

Expand a discovery theme when the startup can serve the customer, the page answers the query, measurement is trustworthy, and activation or pipeline evidence supports more investment. Promote a search term when its product fit and business outcome are clear. Change the page or qualification path when the demand is valuable but the current experience is wrong.

Pause or narrow a theme when the product cannot serve the query, the page promise is inaccurate, the sales or onboarding path cannot handle it, or the account cannot measure the outcome honestly. Record the reason so the same hypothesis is not relaunched without new evidence.

Common startup setup mistakes

Using broad as a substitute for positioning

Broad may discover language, but it cannot decide the startup’s customer, category, offer, or product boundary. Define those before scaling discovery.

Calling every form a success

Forms can include spam, duplicates, poor-fit accounts, and people who never activate. Join acquisition to the next meaningful product or sales event.

Applying exact as a literal-only filter

Google’s current matching behavior uses meaning and context. Exact can be a useful governance role, but search terms and negative rules still need review.

Fragmenting a sparse account

Too many small groups can make every result look inconclusive. Split where a real decision differs; otherwise keep the query job and evidence visible in a coherent theme.

Inheriting another company’s exclusions

A negative list reflects an offer and a market. Recheck every exclusion against the startup’s current product, plans, customers, and goals.

Startup keyword-match setup checklist

Before launch or a match-type change, confirm that:

  1. customer, problem, category, product, competitor, brand, support, and jobs intent are labeled;
  2. each theme has a relevant page, product path, owner, market, and next action;
  3. broad, phrase, and exact roles are documented without promising literal-only traffic;
  4. negatives reflect real fit boundaries and have deliberate scope;
  5. product capabilities, pricing, pages, signup or sales flows, and ads agree;
  6. activation, qualified account, pipeline, paid conversion, retention, or another approved event is measured;
  7. search-term reviews record promote, exclude, move, page, and pause decisions; and
  8. current Google Ads keyword matching guidance has been checked before changing the live account.

The startup audience-targeting guide adds customer and market context, and the Google Ads match-types settings guide covers broader terminology. adz.so can help organize and inspect supported account data, but it does not replace Google Ads controls, search-term review, product analytics, CRM governance, or approval of the startup’s commercial strategy.