Google Ads guide

How to Set Up Google Ads Impression Share for Agencies

To set up Google Ads impression share for an agency, create a repeatable report that keeps each client account, campaign, eligible opportunity, constraint, and business outcome separate. Impression share is the share of estimated eligible impressions that a campaign received, so it is a diagnostic about coverage rather than a universal performance score. The right setup connects that coverage to qualified leads, orders, pipeline, margin, capacity, or another client-approved outcome.

Use Google’s impression-share definition and improvement guidance for platform context. The agency workflow below explains how to make the metric useful across different clients without turning one benchmark into an automatic budget recommendation.

Define the client decision before the report

Start with the decision the report must support. An ecommerce client may be deciding whether to fund a profitable product category. A local client may need more booked calls but have only one technician. A B2B client may want more target-account reach, while a SaaS client may care about qualified trials and activation. These are different reasons to study impression share.

Write down the account’s objective, audience, geography, offer, sales or fulfillment capacity, economics, and primary outcome. Add exclusions such as locations not served, products out of stock, low-margin services, or lead types the client cannot accept. A campaign should not be pushed toward every eligible impression if the business cannot respond to or profit from the added demand.

Give every client a short interpretation note. It can state that the agency is monitoring coverage, budget opportunity, rank opportunity, or deliberate reach. This prevents a client-facing percentage from appearing to be a target that was chosen without context.

Build a client-safe data scope

Set the report scope before adding calculations. Confirm the customer account, manager hierarchy, campaign, date range, time zone, currency, market, network, and conversion actions. Keep the client identifier on every row and use separate filters or views for each account. Never combine spend, impression share, or conversions from unrelated advertisers into one agency benchmark.

Use a consistent base set of fields such as impressions, impression share, clicks, cost, conversions, conversion value, campaign status, and the available lost-impression-share-by-budget or rank fields. Add the fields needed for the client’s model: lead acceptance, call disposition, order margin, branch, product, account tier, activation, or pipeline stage.

If the source data uses different names or windows, document the mapping. A qualified lead counted in the CRM is not necessarily the same event as a platform conversion. The report should preserve both values rather than silently treating them as interchangeable.

Add the useful dimensions

Impression share becomes actionable when the agency can locate the coverage change. Segment the report only by dimensions that answer a real question, such as campaign, service or product group, location, device, search category, or brand and non-brand intent. Too many breakdowns create small, noisy cells that invite overreaction.

For each segment, show the time window and enough volume to judge the signal. A small local campaign can have an unstable daily percentage because a few auctions change the denominator. Use a longer window or a grouped category when a daily view would lead to a false conclusion.

Separate availability from performance. A segment can have low impression share because its targeting is narrow, because demand is limited, because budget is constrained, or because rank is weak. Then compare its qualified result with spend and conversion quality. The question is not merely where share is low; it is where additional eligible coverage would be valuable.

Separate budget from rank

Where Google Ads provides the relevant loss fields, place budget and rank in separate columns. A budget-limited campaign may need a spend model. A rank-limited campaign may need work on bid, relevance, assets, landing page experience, feed quality, or the offer. The fields describe different kinds of constraint and should not produce the same recommendation.

For a possible budget expansion, estimate what the client can afford and serve. Review incremental cost, qualified conversion rate, sales acceptance, margin, close rate, customer value, and operational capacity. A higher share is not a win if it produces inquiries that the sales team cannot follow up or orders that lose money.

For a possible rank improvement, inspect the query or audience context first. Check message fit, landing-page alignment, conversion friction, product availability, location relevance, and the quality of the signal being optimized. Raising bids can buy more auctions, but it cannot repair a broken page or an offer that the target customer does not understand.

Join coverage to the client outcome

Create a second view that joins platform coverage with the client’s accepted outcome. For lead generation, use qualified or accepted leads, booked appointments, response time, and close rate. For ecommerce, use orders, revenue, contribution margin, repeat value, and product availability. For B2B, use account fit, opportunity stage, pipeline, and revenue rather than treating every form fill as equal.

Google’s measurement guidance is useful for understanding platform conversions, but client systems usually explain quality. Preserve attribution rules and time zones, and label whether the report uses click date, conversion date, order date, or CRM update date. If a comparison is not like-for-like, say so.

Look for combinations rather than isolated scores:

  • low share with strong qualified economics may justify an expansion test;
  • low share with weak quality may call for tighter targeting or a better offer;
  • high share with weak outcomes may indicate wasted eligibility or poor qualification; and
  • falling share during a stockout, closure, or capacity limit may be acceptable.

This outcome layer keeps the agency from reporting more exposure as the goal when the client actually needs more profitable or serviceable demand.

Create a repeatable review and approval loop

Give each row or recommendation a clear owner. A useful agency review record includes:

  1. client, account, campaign, market, and date range;
  2. current impression share and the relevant eligibility or loss signal;
  3. the client outcome connected to that segment;
  4. the proposed change, scope, and expected evidence;
  5. the owner, client approver, and review date; and
  6. a stop condition if cost, lead quality, margin, or capacity moves the wrong way.

Use a consistent status such as investigate, protect, expand, improve, narrow, or accept. The status is a decision aid, not a replacement for judgment. Explain what is known, what is inferred, and what data is missing. This makes an approval safer than a slide that shows a red percentage with no operating context.

Run the report on a stable cadence, but do not force every client into the same cadence. A high-volume retailer may support a weekly review; a small local account may need a longer window. Preserve prior snapshots so the agency can explain whether a change came from budget, auction conditions, targeting, demand, tracking, or a client-side constraint.

Avoid common agency setup mistakes

The most damaging mistake is using a shared impression-share target across clients. The metric’s denominator is estimated eligible opportunity, and the client decides whether that opportunity is desirable. A second mistake is treating lost budget and lost rank as synonyms. They lead to different investigations and different owners.

Avoid mixing manager-account totals with client-level conclusions, changing date ranges between the percentage and the outcome, or comparing platform conversions with CRM outcomes without noting the difference. Do not call a low-volume percentage a trend. Do not hide missing rank or budget fields behind a confident recommendation.

Finally, do not promise an adz.so feature that is not live. adz.so can be used as a truthful place to organize and inspect supported account data, while client approval, sales qualification, fulfillment, margin analysis, and revenue evidence remain part of the agency’s operating process.

Agency setup checklist

Before publishing the report, confirm that:

  1. the client objective, business boundary, and outcome are written down;
  2. account, campaign, location, date, currency, and conversion filters are correct;
  3. each row remains attributable to one client account;
  4. impression share is explained as eligible coverage;
  5. budget and rank constraints are separated where available;
  6. coverage is joined to qualified business outcomes and capacity;
  7. recommendations have scope, owner, approver, review date, and stop condition; and
  8. limitations, missing fields, and attribution windows are visible.

The best agency setup is consistent in structure but specific in interpretation. Use the agency impression-share guide for the decision framework, then keep each client’s economics, readiness, and approvals at the center of the recommendation.