Search Query, Keyword & Negative Governance

Every query needs exactly one owner, yet many accounts have never made that decision.

Broad match, Smart Bidding, and years of ad hoc keyword additions can leave the same query matching several keywords while negative lists grow only after waste is noticed. We review the search terms report, assign one owner to each query cluster, and establish the negative and match-type rules needed to maintain that ownership. Accounts running broad or mixed match types with real query volume see the most from this. A small exact-match account where every term is already reviewed by hand rarely needs it.

A query system where you can name which keyword owns any search term, and what stops the ones that shouldn't spend.

A specialist routing a stream of search queries into labeled owner lanes and a tiered negative-keyword filter

Some of the 500+ brands we've worked with

See all references
  • EY
  • Yves Rocher
  • Zorlu PSM
  • Isuzu
  • Hotiç
  • Sportive

How the work runs

Raw queries become a governed system

Each record can be checked against the live search terms report. AI can cluster, flag, and draft. It cannot add a negative keyword, change a match type, or move budget.

How we hold ourselves to it

  • Every query gets exactly one owner
  • Every negative addition is logged with its reason
  • Broad match earns its reach through negative discipline
  • The search terms report gets reviewed on a schedule

  1. Mine and classify the queries

    Pull the search terms report and bucket every query by intent and theme, marking which are covered on purpose and which arrived by accident.

    Classified search-term set with each query bucketed by intent and theme, and intentional coverage separated from accidental.

    AI assist
    Clusters raw queries by intent and flags terms that don't fit an existing bucket.
    Human gate
    Paid-search lead confirms the intent buckets before anything changes.
    Owners
    Paid-search lead + analytics lead

  2. Detect internal competition

    Find every query matching more than one active keyword or ad group, and decide which one should actually own it going forward.

    Internal-competition list naming every query matched by more than one active keyword or ad group.

    AI assist
    Flags every query currently matching multiple active keywords.
    Human gate
    Strategy lead assigns the single owner for each contested query.
    Owners
    Strategy lead + paid-search lead

  3. Build the negative architecture

    Decide what belongs at the account level, since that tier applies automatically across Search, Performance Max, Shopping, and more, what belongs in a shared list, and what stays campaign-specific.

    Tiered negative architecture assigning each exclusion to account, shared-list or campaign level.

    AI assist
    Drafts the tiered negative list from the flagged waste.
    Human gate
    Client owner approves account-level exclusions before they apply broadly.
    Owners
    Client owner + paid-search lead

  4. Set the match-type policy

    Decide, tier by tier, where broad match earns its reach and where phrase or exact match still holds the line.

    Match-type policy stating, tier by tier, where broad, phrase or exact match applies.

    AI assist
    Compares current match types against the proposed policy and flags mismatches.
    Human gate
    Paid-search lead signs off on the policy before it is applied.
    Owners
    Paid-search lead + strategy lead

  5. Apply and route

    Push the negative lists and match-type changes, and record which ad group or campaign now owns each previously contested query.

    Applied negative lists and match-type changes with the new owner recorded for each contested query.

    AI assist
    Drafts the change list with before-and-after state for review.
    Human gate
    A named human applies every live change to keywords and negatives.
    Owners
    Media specialist + paid-search lead

  6. Run the recurring audit

    Set the cadence for reviewing new search terms, refreshing negative coverage, and catching new internal competition before it spends much.

    Audit calendar naming the review cadence, the cycle owner and the open queue carried forward.

    AI assist
    Prepares the review queue from the latest search terms report.
    Human gate
    Paid-search lead reviews and closes each cycle.
    Owners
    Paid-search lead + client owner

What lands with your team

A system that holds up between reviews

The map, register, and policy work together to keep the account from drifting back into the state we found it in.

  • Query-owner map

    Which keyword or ad group owns each query cluster, with the reasoning behind contested calls.

    Accepted when

    Every query cluster names its owning keyword or ad group, and each contested call carries its reasoning.

    Cadence: Built at setup

  • Tiered negative keyword register

    Every negative sorted into account, shared, or campaign level, with its source, owner, and date added.

    Accepted when

    Every negative names its tier, source, owner and date added, so a later removal is a decision rather than a guess.

    Cadence: On add or retire

  • Match-type policy sheet

    Which keyword tiers run broad, phrase, or exact match, and the reasoning behind each tier's policy.

    Accepted when

    Each keyword tier states its match type and the reasoning behind it, not a single account-wide default.

    Cadence: Strategy or evidence change

  • Recurring audit calendar

    The review cadence, the owner for each cycle, and the open queue carried from the last review.

    Accepted when

    The cadence, the owner for each cycle and the queue carried from the last review are all recorded before the cycle starts.

    Cadence: Fixed recurring schedule

Before the work starts

When this work is the right next step

More query reach requires stronger controls. Broad match is built to find more of the query space than phrase or exact match, and Google's own guidance pairs it with Smart Bidding. That reach only pays off if something is filtering what comes with it. Without a negative architecture underneath, the same junk term gets added to one campaign's list and keeps showing up in every other campaign that never got the memo.

A good fit when

  • Broad match widens the query space, but nobody reviews the search terms report on a recurring schedule.
  • The same search query matches multiple active keywords or ad groups, so nobody can explain which one should own that intent going forward.
  • Several people add keywords and negatives across campaigns, yet the account has no tiered register showing the source, level, and owner of each exclusion.
  • The same query matches two or more active keywords, and nobody assigned it there on purpose.
  • Negative keywords get added campaign by campaign, so a term excluded in one place keeps spending in another.
  • Broad match was turned on for the bidding benefit, but the search terms report hasn't been reviewed since.
  • A negative meant to block a whole category only catches the one exact phrase someone happened to type.

Better handled as other work when

  • The account is small and exact-match only, while every visible search term is already reviewed by hand and no ownership conflict appears in the report.
  • Nobody on the team can open the search terms report, so a negative-keyword decision would rest on guesswork instead of classified query evidence.
  • Negative-keyword additions have no approver or recorded owner, so account-level exclusions cannot meet the method's human-approval rule.

People who own a single channel

Paid Search, Paid Social, CRO, and Programmatic each run under a named owner at Zeo. The consultants below are matched to the channel this page is about, so you can see who you'd actually work with.

Content we've produced on this topic

Articleby Metehan Urhan6 Reasons To Focus on Long Tail Keywords

Tools we use

Tools behind this work

  • Adalysis

    mines the search terms report and flags two keywords fighting for the same query

  • Airtable

    the ownership record that says which person is accountable for which query cluster

Keep query conflicts from returning

Give every search query a clear owner

Start with the current search terms report. We map contested queries, set negative tiers, and document the match-type rules that govern the next review.
Clean up your query governance

The questions we hear most

No. Smart Bidding optimizes within the query space it is allowed to compete in. Negative keywords set that boundary, and automation still needs a clean space to work inside, especially under broad match.