01 · Orientation
Brand and non-brand need a boundary you can verify
A low average CPA can conceal a decline in the number of suitable enquiries your business wins from generic searches. If more people are searching for your name at the same time, that less expensive share of demand may improve the account average on its own. Separating brand and non-brand should therefore answer a specific question: which demand does each campaign serve, and how economically does it handle that demand?
Small and medium-sized businesses do not need as many campaigns as possible to achieve this. They need an agreed definition of their own brand, consistent rules for mixed queries, and reporting that keeps unknown shares visible. Campaigns, exclusions and budgets follow from those decisions. Naming a campaign “Non-Brand” does not establish that it receives no searches for your own brand.
Use the account structure map as a shared working record
This guide helps you create an account structure map that records responsibilities, permitted queries, exclusions and evidence from checks. It separates your business definition from what the available controls can do. The Google documentation on brand settings describes the scope of those controls; your own map records which feature is actually configured in the account concerned.
An important distinction: Here, brand means queries referring to your own brand. It does not divide people into existing and new customers. Someone making their first enquiry may search for your name, while a longstanding customer may arrive through a generic product term.
Technical information checked on 15 September 2026. The examples and figures are constructed for illustration and do not describe results from a client project.
02 · Definition
First define what belongs to your own brand
Start with a brand register. Record the company name, shortened names people actually use, the domain name and distinct product brands. For each entry, check whether searchers clearly mean your business. A descriptive company name such as “Büroausstattung Direkt” can also express a generic need for office equipment. Such cases require a reasoned decision, rather than an automatically imported list of words.
The fictional business Falkenwerk Büroeinrichtung sells furnishings for commercial workplaces. “Falkenwerk” and “Falkenwerk Büroeinrichtung” refer to its own brand. “Büromöbel Frankfurt” is generic. A manufacturer's brand stocked by the business does not become its own corporate brand merely because it appears in the range. A competitor's name also receives a separate classification.
Your brand register and Google's brand list serve different purposes
Your register contains the business definition, its rationale and the person responsible. A brand list in Google Ads is a technical object used by supported brand settings. If the appropriate brand is missing, you can request its addition. The brand URL displayed in the interface distinguishes brands with similar names; it does not restrict countries or search queries. Check the actual entry, alternative names and distinct subsidiary brands before attaching the list to campaigns.
A single domain can represent several names, while one brand may appear on several domains. Maintain more than a list of URLs. Also record former names that people still search for and ambiguities that remain unresolved. Keep a term in the review category until it can be classified with sufficient confidence.
STOP: A comprehensive exclusion list is not ready for approval until your own brand has been distinguished from retailer brands, competitors and generic terms.
If you manage brands belonging to several related businesses, define the relationship separately for each account. A sister company's brand is not automatically the own brand of the entity being advertised. The same principle applies to retailers and manufacturers. Record whether a shared umbrella brand is explicitly included. This prevents a technically correct brand-list selection from assigning demand to the wrong business.
03 · Search queries
Assign each query to one unambiguous category
Use mutually exclusive categories in the operational report. Additional analytical labels are useful, but they must not cause the same query to enter the total twice. Apply an agreed order to mixed queries: first combinations of your own brand and a competitor, then other references to your own brand, followed by competitors, third-party product brands, generic demand and unresolved cases.
Seven categories for Falkenwerk Büroeinrichtung
| Category | Example | Own brand | Intent | Responsibility | Check |
|---|---|---|---|---|---|
| Own brand only | falkenwerk | Yes | Find the business | Brand Search | Unambiguous reference to the name |
| Own brand and product | falkenwerk schreibtisch | Yes | Explore the offer | Brand Search | The product reference matches the business's offer |
| Own brand and service | falkenwerk lieferstatus | Yes | Resolve an existing transaction | Separate service decision | Do not automatically count as sales demand |
| Own brand and competitor | falkenwerk oder nordkontor | Yes, mixed | Compare suppliers | Separate comparison category | Count once and assign responsibility |
| Competitor only | nordkontor büromöbel | No | Find another supplier | Separate decision | Exclude from the core generic category |
| Third-party product brand | markenstuhl herstellername | No | Find a specific manufacturer's product | Product-range group | Distinguish manufacturer from retailer |
| Generic or unresolved | büromöbel frankfurt | No or unresolved | Express a need without the business's name | Non-brand or review | Label unresolved cases separately |
The final row describes a decision branch, rather than a combined reporting metric: clearly generic queries enter non-brand, while unresolved queries remain separate in the report. For mixed comparison queries, the business must explicitly decide whether paid advertising is wanted and where it should run. Assigning them to a brand campaign does not remove their separate reporting category.
04 · Account map
Map responsibilities before creating campaigns
The account structure map gives each query category an operational owner. It records which campaign is intended to receive the demand and the evidence supporting that assignment. Mark optional inventory explicitly. For example, a service business without a product feed does not need a Shopping campaign simply to make the map look complete.
A compact structure with five possible areas
| ID | Purpose | Permitted demand | Boundary | Goal and budget | Evidence |
|---|---|---|---|---|---|
| S-B01 | Brand Search | Approved own-brand categories | Explicit rules for service and comparison queries | Qualified lead; separate budget decision | Settings and visible search queries |
| S-N01 | Non-Brand Search | Relevant generic demand | Exclude own brand at campaign level | The same qualified lead; separate budget decision | Exclusions, goal and search-query review |
| S-X01 | Special cases, where justified | Approved comparisons or third-party brands | Separate categories even under shared management | Justified use and a defined risk limit | Approval and economic assessment |
| P-01 | Performance Max, optional | Defined purpose across supported channels | Brand exclusion with documented scope | Appropriate goal; separate budget decision | Search, channel and settings review |
| SH-01 | Shopping, optional | Products from a suitable feed | Check brand references and product logic | Commercially appropriate product goal | Feed, search queries and exclusions |
In this example, S-B01 and S-N01 use the same definition of a qualified lead. Their bidding decisions can still differ. Keeping those issues separate prevents an apparent improvement from arising solely because one campaign uses an easier conversion goal. The IDs are examples, rather than a required naming convention.
05 · Scope
Separate campaigns where a distinct decision is needed
A separate campaign makes sense when budget, bidding approach, audience, geographic scope or responsibility requires independent management. Wanting another row in a report is not always enough. You can also create reporting categories within a manageable campaign structure. Every additional campaign brings maintenance work and may leave fewer observations within each decision-making area.
For Falkenwerk, a dedicated brand campaign may be justified if the business wants to limit spending on known demand clearly and assess generic demand independently. A separate campaign for every spelling of its name would duplicate the same purpose. Those variants belong first in the brand register and in the controls used to cover the brand.
One decision and one accountable area
Before any split, ask three questions: which decision will improve, who will make it, and what data will be sufficient? If there is no defensible answer, a reporting label will often be enough. This matters especially where budgets are small and qualified enquiries are infrequent. Adding campaigns does not automatically make a weak evidence base more precise.
Also record whether a structural change only reallocates responsibilities or changes reach as well. If exclusions, bids and target regions change at the same time, it becomes difficult to connect a later outcome with a particular cause. Plan those interventions separately wherever operations allow.
Approval rule: Every new campaign needs its own purpose, an accountable owner and a boundary that can be checked. A larger account map is not a measure of quality.
Multiple languages or regions do not justify automatic duplication either. Check whether the offer, budget ownership and customer handling really differ. If areas are managed separately, treat the same brand consistently while recording local ambiguities. A German brand term can mean something different in another language. The central taxonomy should document such exceptions without forcing them into an apparently uniform definition.
06 · Baseline data
Establish a baseline with comparable goals
Before making a change, export a sufficiently mature period containing costs, conversions, conversion values and visible search queries. Add the conversion goals actually in use and the settings of the affected campaigns. A screenshot showing a campaign name is not enough. What matters is which actions count towards optimisation and which data arrives after a delay.
Choose the period according to demand, seasonality and your actual sales cycle. A standard minimum duration is no substitute for checking these conditions. If the CRM only confirms qualified enquiries later, the periods being compared need similar data maturity. A recent period with outstanding sales feedback can initially make CPA look artificially worse.
Document search intent alongside measurement
SaleStudia's supporting guide to keyword research and search intentcan help you classify generic and brand-related terms. Use only criteria that fit your offer. Retain the original query alongside its category so that a later correction remains traceable.
Also note public holidays, promotions, form changes and major interruptions to sales operations. Such events can affect brand and non-brand differently. Your baseline should therefore include figures and a short timeline of relevant events. It provides a basis for comparison, rather than proof of what the account would have done without the change.
07 · Search
Check brand inclusions and exclusions in the actual account
New brand inclusions and brand exclusions in Search require AI Max at campaign level. Existing settings may continue to work outside AI Max: you can still add brands to existing lists and remove settings. Under the documented transition, attaching a new brand list requires AI Max. Check the setup path available in your account rather than following an old guide unchanged.
Configuring a technical boundary for brand demand is a separate decision from adopting AI Max as a whole. SaleStudia's supporting guide to negative keywords and irrelevant search queries helps explain the alternative exclusion approach. Negative keywords do not fully reproduce brand recognition. Before activation, record which features you actually need and which other settings will become relevant.
Ad group rules do not automatically form an intersection
Check the Google instructions for brand inclusions in Search for the level concerned. When a brand list is configured at ad group level, it overrides the campaign-wide inclusion rule for that group. Do not assume that both lists automatically apply as an intersection.
Conflict rule: A brand exclusion takes precedence over a conflicting inclusion. Negative keywords can also block brand searches you want to reach. Review every effective level together before approving a campaign.
Save the approved lists, levels and exceptions in the account map with a date. A label such as “brand allowed” cannot be checked reliably later without those details.
08 · Performance Max
Describe the scope of PMax exclusions precisely
Performance Max includes several channels and formats. A brand exclusion therefore does not automatically make the entire campaign completely free of brand demand. Google's dedicated instructions mention Search, Shopping and YouTube Search. Its general overview still mentions only Search and Shopping. Record the confirmed scope and the date of the review. The dedicated instructions do not establish an exclusion across all YouTube, video or Display placements.
Also check whether a Shopping exception is available. If Shopping ads are allowed to appear for searches involving excluded brands despite the brand exclusion, an intentional overlap remains in that inventory. This may be a justified product decision. However, the report must then explain the exception instead of describing the campaign as fully separated from searches for your own brand.
Make exceptions visible in the account map
The Google instructions for brand exclusions in Performance Max provide the basis for this check. Record the excluded brands, the confirmed inventory scope and the state of any available Shopping exception. Recheck the actual settings after changes.
Do not treat PMax as a universal replacement for Brand Search or assume that it is a purely non-brand channel. Your account map can explicitly give it a mixed role. Budget owners need to understand that scope and avoid equating its results with those of a more narrowly defined Search campaign. A high PMax ROAS does not, by itself, identify the category of demand being served.
If Shopping is deliberately intended to receive own-brand queries, record why. There might, for example, be a distinct product objective that will then undergo an economic assessment. That decision does not establish that the additional exposure is automatically necessary or profitable. An exception needs the same accountable review as an exclusion. It should not remain simply because it was enabled unnoticed during the original setup.
09 · Exclusions
Avoid accidentally blocking your own brand campaign
Negative keywords complement brand controls but work differently. Document their scope: account, shared list, campaign or ad group. The account-level negative keywords apply automatically to relevant Search and Shopping inventory in eligible campaigns. Excluding your own brand there can also block the brand advertising you explicitly want to run. An exclusion intended to clean up non-brand traffic should therefore not be added at account level without checking its consequences.
Google's guidance on negative keywordsexplains which spelling variations are handled automatically. According to the current description, capitalisation differences and misspellings are covered automatically. Add synonyms and singular or plural forms where needed; there is no general expansion to close variants. Negative broad match blocks queries containing all of the negative keyword's words in any order; phrase match requires the same order, and exact match allows no additional words. A negative term therefore does not block every query with a related meaning.
Maintain lists as documented rules
Check each list against specific queries that you want to allow and block. An overly broad exclusion can suppress valuable generic demand, especially where the brand name contains an ordinary descriptive term. Record the rationale, owner and date for every addition. Older entries should still be explainable later.
The negative keywords in Performance Max apply only to Search and Shopping inventory. Existing keyword lists can also be used at campaign level. Such a list does not replace a separate assessment of other placements. Keep brand lists and keyword lists distinct so that similar names do not lead someone to assume that they work in the same way.
STOP: If an own-brand term is blocked at account level, investigate and correct that conflict first. Raising bids does not overcome an effective exclusion.
10 · Selection
Do not confuse prioritisation with guaranteed routing
When several campaigns could potentially reach the same query, selection and prioritisation rules apply. In the documented comparison, an eligible exact-match keyword in Search that is identical to the query is prioritised over phrase or broad match, search themes and Performance Max. This prioritisation concerns Search ads; it does not create equivalent precedence over Shopping ads. Nor does it guarantee that every brand-related query will always reach the intended brand campaign. The actual conditions for serving must be met.
The Google rules for prioritising ad groups and asset groups distinguish identical matches, eligibility and other selection criteria. Check the intended campaign's budget, approvals, targeting and other restrictions. A campaign that cannot participate in the particular selection does not gain priority from its name.
Check positive coverage and exclusion boundaries together
Include the important brand forms in the responsible Search structure and check the exclusions in other areas. These two sides serve different purposes. An exact-match keyword also cannot represent every possible mixed query as an identical string. Search queries and the effective settings remain relevant to the actual distribution.
Individual searches that you perform yourself do not provide complete evidence. Auction conditions, location, time and other circumstances can differ. Use appropriate preview and diagnostic tools alongside the available reports. Where distribution differs from expectations, investigate eligibility and conflicting rules before splitting the structure further.
11 · Money and goals
Separate budget decisions from the conversion definition
Brand and non-brand can have different commercial roles without “qualified lead” meaning something different in each area. In this example, S-B01 and S-N01 both use the same enquiry confirmed as qualified by sales. Different target CPAs or other appropriate bidding decisions need a rationale tied to each area's purpose.
Check the conversion actions actually used for bidding. A common goal label is insufficient if different actions contribute to optimisation. In particular, record whether custom goals are used and which actions they contain. The shared definition should make comparison possible without hiding the different roles of the two areas.
Show shared budgets and portfolio connections
Separate campaigns do not automatically provide independent budget control. The link between shared budgets and portfolio bid strategies can keep the areas connected. With this link, the budget group follows the campaigns belonging to the portfolio strategy. Record those relationships explicitly in the account map. If decisions need to be independent, the effective settings must support that requirement.
Do not adopt a fixed split such as ten per cent brand and ninety per cent non-brand as a universal rule. Requirements depend on brand awareness, competition, the offer and economic constraints, among other factors. Decide according to the role and mature results. A low brand CPA does not justify unlimited additional spending, and a higher non-brand CPA is not automatically uneconomic.
Alongside CPA, compare the absolute number of qualified enquiries and their economic value where reliable information is available. An area may become cheaper because it reaches only a few straightforward cases. Conversely, a higher CPA may be associated with more valuable orders. Without that context, even correctly separated brand reporting is too narrow to support a sound budget decision.
12 · Worked example
Recognise when a better average conceals a problem
The following simplified calculation contains only brand and non-brand. It deliberately excludes other categories, unknown shares and changes to measurement. Every conversion represents the same type of qualified enquiry. CPA is cost divided by conversions in each row; the overall figure is calculated from the combined totals.
More brand conversions can improve the overall CPA
| Period | Area | Cost | Qualified leads | CPA | Interpretation |
|---|---|---|---|---|---|
| A | Brand | €1,000 | 100 | €10.00 | Low-cost known demand |
| A | Non-brand | €9,000 | 90 | €100.00 | Baseline for generic demand |
| A | Total | €10,000 | 190 | €52.63 | Weighted account figure |
| B | Brand | €2,000 | 200 | €10.00 | Larger low-cost share |
| B | Non-brand | €8,000 | 64 | €125.00 | Fewer leads and higher CPA |
| B | Total | €10,000 | 264 | €37.88 | Better average despite worse non-brand performance |
The calculation establishes neither a cause nor the optimal budget allocation. It illustrates a change in composition. In a real account, also examine demand, competition, values and data maturity. Different reporting periods and outstanding sales feedback can introduce further distortion into this type of comparison.
Report cost, results and CPA for each category alongside the overall figure. Do not calculate overall CPA as the simple average of the two category CPAs. This keeps visible which share supports the improvement and where a separate investigation is needed.
13 · Visibility
Give unknown search queries their own place
A search terms report does not necessarily show every individual query. Google explains the available information and its limitations in the search terms report documentation. Visible rows therefore do not represent the entire population of searches. Always state the basis used to calculate a percentage in your report.
Consider a simplified example with fully comparable Search inventory and reporting periods. The Search campaign being examined has 1,000 clicks, of which 800 can be matched to visible queries. Of those, 80 refer to the business's own brand. The observed brand share is 80 divided by 800, or ten per cent of the clicks that can be classified. The query category remains unknown for the other 200 clicks.
Unknown means neither brand nor non-brand
Do not apply that ten per cent to all 1,000 clicks without further evidence. Equally, the 200 unknown clicks cannot automatically be recorded as generic demand. Report the classification coverage of 800 out of 1,000 clicks separately from the brand share within the visible set.
A cost share requires the costs of the corresponding visible, classified rows; a click share is not a cost share. Use the same metric, period and scope for numerator and denominator. If reports provide different aggregations, document that limitation instead of filling the gap through calculation.
14 · Multiple reports
Read channel and brand information separately
Search is a channel; brand is a classification of the query. The two dimensions answer different questions. The Search share of Performance Max does not automatically reveal how much own-brand demand it includes. Conversely, a brand share calculated from visible search terms does not describe the campaign's other channels.
The Google documentation on the Performance Max search terms report describes breakdowns including ad format and landing page. This view supports checks of Shopping exceptions; it is not a complete channel report. Data is available from March 2023, and Store visits and Store sales conversions are not included. Record those limitations alongside your query classifications.
Ask one specific question of each report
Use the channel report to ask where serving and measured results occur. Use the search terms report to identify visible queries relating to your own brand. Use the CRM to establish which enquiries were actually qualified. Another perspective comes from the paid and organic report after linking Search Console. It covers paid text ads, not Shopping ads; organic data begins when the link is established and does not follow the same geographic and language settings as Ads. Combined click figures therefore do not establish incremental advertising impact either.
Even a high average ROAS for a channel does not demonstrate that the next additional euro invested there will earn the same return. A budget change needs its own hypothesis and observation. The account map should make these limitations visible so that different metrics do not merge into an apparently conclusive recommendation.
15 · Implementation
Introduce changes with clear stopping rules
Before the transition, prepare a short change plan covering the baseline, affected campaigns, new rules, owner and review date. Begin with a manageable part of the account. After setup, first check technical conflicts and serving, then assess mature results. Early operational checks are useful; premature economic conclusions are not.
Define in advance which deviations trigger investigation or a stop. Examples include a blocked brand campaign, an unintended account-level exclusion, a different conversion goal or an unapproved Shopping exception. Economic limits should fit the business. Do not move them retrospectively until every outcome can be presented as a success.
Distinguish structural maintenance from an effectiveness test
A before-and-after comparison alone does not isolate the change's effect. If you want stronger evidence about a specific effect, use SaleStudia's guide to planning and evaluating Google Ads experiments to establish a clear hypothesis and decision rules in advance. Not every account restructuring can be represented meaningfully by every experiment type.
Also document which original settings can be restored if needed. Reverting settings does not restore earlier auction conditions or results. Complete the technical implementation only after responsibilities and exclusions have been checked; observation of the economic outcome may continue afterwards.
Agree on a concrete sequence for the first review: check effective settings, classify unusual queries, reconcile goal actions and investigate unexpected cost developments. Assign an owner to every deviation found. Avoid making several corrections at once without recording them, because this obscures both the original error and the effect of its repair. After a necessary immediate correction, restart the documented observation from the changed state.
16 · Working register
Complete the account structure map with evidence
Use the following register for every significant boundary in your structure. The blank fields are intended for your own entries. Record the current state and its evidence separately, so that “configured” is not confused with “verified”. An owner and a date make unresolved questions actionable.
Eight areas to check before approval
| Review area | Record | Evidence | Owner and date | Status or STOP |
|---|---|---|---|---|
| Brand definition | Own names, variants and unresolved cases: ___ | Register and reasoned classification: ___ | ___ | ___ |
| Query categories | Classification order and mixed cases: ___ | Reviewed example queries: ___ | ___ | ___ |
| Campaign responsibility | ID and permitted demand: ___ | Approved account map: ___ | ___ | ___ |
| Search rules | Levels, lists and AI Max status: ___ | Settings and conflict check: ___ | ___ | ___ |
| PMax and Shopping | Scope and approved exceptions: ___ | Settings and reporting limitations: ___ | ___ | ___ |
| Goals and money | Conversion definition, budgets and portfolios: ___ | Effective goals and connections: ___ | ___ | ___ |
| Measurement basis | Period, maturity, numerator and denominator: ___ | Export, CRM reconciliation and unknown share: ___ | ___ | ___ |
| Decision | Change, limits and next review date: ___ | Approval and its rationale: ___ | ___ | ___ |
A well-maintained map does not establish incremental advertising impact. Whether advertising produces additional results compared with a suitable control condition requires an appropriate effectiveness study. Google's Conversion Lift measurement describes one such approach, subject to its availability requirements.
Retain older approved versions. This makes it possible to distinguish a changed classification or technical configuration from a real change in performance when a metric looks different later.
17 · Common questions
FAQ on separating brand and non-brand
Does every small business need its own brand campaign?
No. The question is whether budget, bidding or responsibility needs an independent decision and whether the benefit justifies the maintenance work. Even without an extra campaign, you should be able to identify brand references in visible queries. A campaign is a unit of control, rather than a prerequisite for every reporting category.
Do brand conversions automatically represent existing customers?
No. The query describes a brand reference, not the person's customer history. A prospect may search for your name for the first time after a recommendation. An existing customer may return through a generic query. New and existing customers require a separate definition and appropriate data for identification.
Is an exact-match keyword for the company name enough?
It helps provide deliberate coverage and can receive priority under the documented conditions. It does not guarantee complete allocation of every brand variant or mixed query. Also check eligibility, budget, exclusions, brand lists and the available search terms reports.
Can I simply exclude my own brand at account level?
That is inappropriate if you want to advertise the same brand in a brand campaign. An account-level exclusion can block the very advertising you intend to run. Set the boundary at the appropriate level, and check shared lists and other effective exclusions for conflicts.
Is Performance Max entirely non-brand after a brand exclusion?
That statement is too broad. The exclusion applies to the documented inventory scope, and an available Shopping exception may deliberately permit brand references. Other channels are not automatically filtered in the same way. Describe the purpose, scope and remaining limitations of the particular campaign.
Which percentage reliably shows unwanted brand demand?
A percentage is interpretable only when its denominator is clear. For example, divide own-brand clicks from visible queries by all corresponding clicks that can be classified. Report unknown shares separately. Cost, clicks and conversions are different measures and must not be mixed within the same ratio.
When is a comparison reliable after a transition?
When definitions and measurement are comparable, relevant delays have been allowed for, and enough suitable observations are available. There is no universal number of days. Also record seasonality, offer changes and other influences. Without an appropriate control, a before-and-after comparison remains limited in the causal conclusions it supports.
Does a low brand CPA establish incremental impact from brand advertising?
No. CPA describes the attributed cost per measured conversion within the chosen measurement approach. It does not reveal which outcomes would have occurred without advertising. A conclusion about incremental impact requires a suitable control condition and an appropriate study design.
18 · Decision
Approve the structure after a shared review
A defensible separation starts with the brand register and ends with an understandable decision. The account map connects those two ends: it shows where each query category belongs, which technical boundaries actually apply and which information the report lacks. Management, marketing and sales can then use the same figures with the same meaning.
The next step is a completed map
Start with S-B01 and S-N01 if your business needs both areas. Add special cases and other campaign types only when they have a specific purpose. Before approval, check the shared lead definition, every exclusion level, budget connections and the documented unknowns. Record the decision in writing.
Would you like to review your account boundaries clearly? SaleStudia supports the analysis and management of Google Ads campaigns for businesses. The starting points are your actual account structure, measurement basis and commercial objectives.