maitiq guide
How do you compare Google Ads agencies fairly?
maitiq · Published
Compare Google Ads agencies on a uniform services and governance grid. Price, certificates and presentation are not enough. What counts is measurement, scope of services, access to the account, decision rights, evidence and handover capability.
Define the shared brief before the tender
A comparison is only as good as the shared starting point. Write down accounts, markets, languages, campaign types, media budget scenarios, the primary conversion, internal roles and the operating model you want. Mark anything unknown explicitly as unknown instead of estimating it. Without the same brief, agencies answer different questions even when their offers look similar.
For an overview of costs, see the agency cost guide. This worksheet starts where a concrete offer has to be made comparable.
The ten criteria of a comparable offer
The following ten criteria map one-to-one onto the rows of the comparison matrix further down. Five of them are binding checks: measurement, a precise service scope with precise objectives, access to the account, approvals and change control, and comparable costs defined for the same scope. Neither reporting nor a discount will later make up for a missing requirement in any of these five areas. That exactly these five checks are binding is a rule of this worksheet, not a general market standard. Assess maitiq against exactly the same ten criteria.
1. Which conversion is optimised? (Knock-out criterion)
The offer must name a primary conversion action and the business context. “More leads” is too imprecise as long as validity, qualification and duplicates remain open. Google distinguishes conversion actions from the goals that group them: when a campaign uses the standard goal of a primary conversion action, that action appears in the Conversions column and can influence bidding. Secondary actions normally appear only under All conversions, but if a campaign includes them through a custom goal, they can affect bidding and reporting regardless of their secondary status. The counting option “every” or “one” sets how conversions after an ad interaction are counted – it does not mean that every click creates a conversion. Two offers can therefore report different lead numbers for identical activity. Ask which event is optimised, which data source applies, what is invalid or duplicate, and who owns the measurement definition; the complete review chain is described in the conversion tracking audit.
2. Is the scope of services described exactly? (Knock-out criterion)
Ask every provider for the same services table: accounts, countries, languages, campaign types, search term and exclusion review, budget and bid control, tracking responsibility, ads, assets, landing page or feed work, reporting, review, response time and projects outside the retainer. Require an “included”, “at an extra charge” or “not included” for every row. Without that mapping, “ongoing optimisation” and “full service” receive no rating.
3. Who owns the account, the data and the payment profile? (Knock-out criterion)
Your business should keep direct administrator access. Google describes five access levels – email only, billing, read only, standard and administrator – and only the administrator level may grant or change access. A client account has at most one owning manager account at any one time; ownership rights also apply to higher-level manager accounts. The client account keeps its data, and an administrator on the client side can unlink an owning manager account. Clarify in the contract as well who holds the payment profile, whether Google bills you directly or the provider passes on the media budget, and who owns Analytics, tag, Merchant Center and feed access.
How maitiq fares: the entry point is a separately arranged, read-only audit. The assessment runs on the agreed connected account or on separately supplied native reports; file handover is arranged separately. The coverage states are complete, limited or skipped. Administrator access stays with you.
4. Who may approve changes and budgets? (Knock-out criterion)
The offer must separate strategy, recommendation, approval and implementation. Record the permitted change types, monetary limits, automation rules, review by a second person, emergency powers, audit trail and rollback. A general authority does not replace limits, and a limit without a stop rule is not a control.
How maitiq fares: analysis, decision and implementation remain separate, logged steps. If the limited autopilot is explicitly activated, it can accept suitable proposals within the configured limits and apply them through the controlled implementation step; the permissions and guard rules of the respective proposal type still apply, and some proposals need a human decision. Without that autopilot, a proposal is implemented after your express approval through a separately authorised and logged step. An express, documented decision to exclude exactly the listed keywords can already authorise that exclusion. In the audit, the autopilot is suppressed; audit access stays read-only and changes nothing in the account.
5. Who actually works on the account?
Ask for named roles rather than titles: who steers operations, who reviews, who escalates, who stands in, how many accounts these people handle in parallel, and what applies during holidays, resignation or illness. A sales presentation with senior advice does not prove who will run the account later. Record whether a change of personnel is announced and whether you may reject replacements.
6. How often are decisions taken?
The review rhythm describes more than a calendar appointment. It should define the necessary documents, prepared decisions, responsible people, minutes, open items and follow-up. Ask which decisions may be taken between two meetings without checking back, and how short-notice topics are handled without bypassing the agreed process.
How maitiq fares: in the managed model, a named Client Success Manager is your fixed point of contact, and the recurring account analysis runs daily by default as a workflow – budget pacing, target CPA adjustment, search term and exclusion maintenance, ad review. The review meeting discusses prepared proposals and records decisions and next steps.
7. How does the fee behave when things change? (Knock-out criterion)
Have the media budget, setup, management, variable fees, tools, creative work, internal dependencies, and switching and termination costs shown separately. What matters is the behaviour, not the initial figure: what happens with a halved budget, a doubled budget, an additional market or a paused campaign? With percentage models, ask for the rate, the basis, the minimum fee, the tiers and the cap. Then convert all models to the same period with the total cost calculator.
How maitiq fares: the offer applies per account for a defined scope of services. The fee is not tied to the media budget – maitiq does not earn more when your advertising spend rises.
8. What evidence do you receive?
A dashboard alone does not demonstrate good account management. In its change history, Google logs who changed what and when for around two years, but it does not list every account-level setting and not every change made by Google representatives; according to Google, the undo function covers only the last 30 days. An existing log therefore proves activity, not judgement. Ask in addition for a rationale for every material recommendation, the data source and period, known gaps, discarded alternatives and the result after implementation. Missing measurement must never be interpreted as success.
How maitiq fares: every proposal carries the measure, the rationale and the supporting data with the old and the new value where such a comparison of values exists; the decision and the implementation are logged. The record shows not only the change but also the thinking behind it – exactly the gap that manual changes without a documented rationale leave behind.
9. Are reporting and documentation defined?
Check whether a report separates media costs, management costs and business results. Ask about the data freshness, time zone, currency, attribution model and how corrections are handled. Clarify which documents are maintained continuously – the measurement concept, naming conventions, automations, scripts, audiences – and whether that documentation belongs to you. A visualised number without a definition remains unsuitable for a contract decision.
How maitiq fares: separately supplied CRM and offline data can be prepared and checked as part of agreed work; a live CRM upload to Google Ads is currently not available through maitiq. The read-only audit makes no uploads and changes nothing in the account. Media costs, fees and business outcomes remain separate figures; evaluating business outcomes or fees requires the relevant supplied data or offer.
10. Is the exit manageable?
A robust offer also describes the offboarding: handover of access, an inventory of accounts and links, active automations, scripts, tracking documentation, open tests, budget and bid status, recent decisions, and data return or deletion. Have termination and cooperation deadlines put in writing. For the switch itself, use the switching and handover guide. No representative Swiss standard for deadlines, fees or response times is substantiated; state your own requirement instead of an alleged market standard.
Comparison matrix with your own weights
Five checks are necessary in this grid: measurement, a precise service scope with precise objectives, access to the account, approvals and change control, and comparable costs. For each check you choose a state: “Met” counts 2, “Partial” 1, “Not met” 0; “Unknown” is not the same as “not met”, is left out of the weighted fulfilment and stays in the coverage with its weight. Partial is not met. As long as a necessary check is not confirmed as met, both values remain unavailable; the offer counts as not yet comparable.
Your team then assigns its own weight from 0 to 5 per criterion; the five necessary checks need at least 1, because a 0 must not remove them from the calculation. All weight fields must be filled in: a blank field means “unknown” and blocks the display of results instead of counting as zero. You rate each provider with the four named states:
Met(2): concrete, accountable and verifiably evidenced;Partial(1): addressed but not evidenced as met;Not met(0): the requirement is answered and not met;Unknown: not clarified; does not count towards the weighted fulfilment, remains in the coverage with its weight and is explicitly not the same asNot met.
weighted point = your own weight × state value (Met 2, Partial 1, Not met 0)
weighted fulfilment = sum of the weighted points ÷ (2 × covered weight) × 100
coverage = covered weight ÷ total weight × 100A criterion rated Unknown is left out of the weighted fulfilment: it counts neither in its numerator nor in its denominator. Its known weight, however, remains in the total weight and therefore in the denominator of the coverage, so the coverage falls. An optional criterion with weight 0 is deliberately excluded from the calculation. All weights must be filled in; a blank field means “unknown” and blocks the display of results instead of counting as zero. Always report both values: the coverage and the weighted fulfilment, for example “72% weighted fulfilment at 80% coverage”. That way an unresolved question is counted neither silently as zero nor as a top score.
Compare offers with your own weights
All fields start blank; no provider is prefilled and no advantage is built in. "Unknown" is not the same as "not met": an unconfirmed mandatory check makes the offer not scorable instead of marking it down.
| Criterion | Weight 0–5 | Offer A | Offer B |
|---|---|---|---|
| Primary conversion and measurement methodMandatory | Mandatory criterion: weight 1 to 5. | Offer A: Primary conversion and measurement method | Offer B: Primary conversion and measurement method |
| Exact service scope and objectivesMandatory | Mandatory criterion: weight 1 to 5. | Offer A: Exact service scope and objectives | Offer B: Exact service scope and objectives |
| Access levels and roles documentedMandatory | Mandatory criterion: weight 1 to 5. | Offer A: Access levels and roles documented | Offer B: Access levels and roles documented |
| Approvals and change controlMandatory | Mandatory criterion: weight 1 to 5. | Offer A: Approvals and change control | Offer B: Approvals and change control |
| Named people, roles and cover | 0 excludes this optional criterion from the calculation. | Offer A: Named people, roles and cover | Offer B: Named people, roles and cover |
| Review and decision schedule | 0 excludes this optional criterion from the calculation. | Offer A: Review and decision schedule | Offer B: Review and decision schedule |
| Costs and service scope clearly definedMandatory | Mandatory criterion: weight 1 to 5. | Offer A: Costs and service scope clearly defined | Offer B: Costs and service scope clearly defined |
| Evidence, references and review record | 0 excludes this optional criterion from the calculation. | Offer A: Evidence, references and review record | Offer B: Evidence, references and review record |
| Reporting and documentation | 0 excludes this optional criterion from the calculation. | Offer A: Reporting and documentation | Offer B: Reporting and documentation |
| Handover, data export and contract end | 0 excludes this optional criterion from the calculation. | Offer A: Handover, data export and contract end | Offer B: Handover, data export and contract end |
Offer A
Not scorable: Primary conversion and measurement method: weight missing; Exact service scope and objectives: weight missing; Access levels and roles documented: weight missing; Approvals and change control: weight missing; Named people, roles and cover: weight missing; Review and decision schedule: weight missing; Costs and service scope clearly defined: weight missing; Evidence, references and review record: weight missing; Reporting and documentation: weight missing; Handover, data export and contract end: weight missing.
Offer B
Not scorable: Primary conversion and measurement method: weight missing; Exact service scope and objectives: weight missing; Access levels and roles documented: weight missing; Approvals and change control: weight missing; Named people, roles and cover: weight missing; Review and decision schedule: weight missing; Costs and service scope clearly defined: weight missing; Evidence, references and review record: weight missing; Reporting and documentation: weight missing; Handover, data export and contract end: weight missing.
Weighted fulfilment = weighted points / (2 × covered weight) × 100. Coverage shows how much weight could actually be assessed. This result does not guarantee service quality or performance and does not determine which provider you should choose.
All weight fields must be filled in; a 0 excludes an optional criterion from the calculation. Both offers stay N/A while weights are missing.
The matrix has the fixed column headings “Offer A” and “Offer B”. Assign Offer A and Offer B to your chosen providers; maitiq may be either one, with no bonus built in. You can record provider names in your own notes or printout. The degree of fulfilment is not a scientific quality grade and not a general ranking; it only shows how completely an offer answers the requirements you defined.
How maitiq fares in this grid
maitiq is not self-service software and not a one-off engagement, but a managed Google Ads service operated by Noomera GmbH in Switzerland. By its own account, maitiq positions its management fee at half that of classic agency support for a comparable management scope; the media budget is excluded. This is maitiq’s own claim, not independently measured customer proof and not a guarantee.
For your matrix this means: the entry point is a read-only audit that assesses the measurement and the access to the account on the basis of your real account. You substantiate the scope of services, contractual responsibility, fee mechanics and decision rights with the offer and the relevant documents; the audit cannot replace them. The factors that determine the quote are on the maitiq pricing page.
How do you run a comparable offer process?
Step 1 – shared briefing: Send all providers the same starting point and separate mandatory requirements, options and unknowns. Require that assumptions remain visible in the offer.
Step 2 – written response: Use the same fields for scope of services, people, rhythm, measurement, access, fees, approvals, documentation and exit. Free text may explain an answer but must not replace a mandatory field.
Step 3 – clarification: Ask every provider the same core questions. Additional follow-up questions are allowed but are assigned in the record to the corresponding criterion.
Step 4 – evidence review: Do not ask for another client’s data. An anonymised example of decision documentation shows the form without passing off a client’s results as proof for your account. Certificates prove a qualification or partnership, not automatically scope of services, control quality or commercial impact.
Step 5 – commercial comparison: Convert all fees to the same period and the same budget scenarios. Mark additional costs and items that are still unknown separately.
Step 6 – accountable decision: Record the status of the knock-out criteria, the chosen weights, the ratings, open risks, decision-makers and the rationale. A points total alone is not a rationale.
Which answers require clarification?
- “We optimise continuously” without a rhythm, responsible people or evidence.
- “Full service” without included and excluded services.
- a total price without separating media, management and additional services.
- account, payment or tracking access only after the contract starts, without a role plan.
- performance-based pay without a data source, attribution and corrections.
- budget autonomy without a monetary limit, stop rule and audit trail.
- offboarding “by arrangement” without a data, tool and access inventory.
These formulations do not prove that an agency is bad. They are incomplete answers that must be made concrete before a decision. Ask every provider how it will substantiate them later: at maitiq, every proposal carries the measure, the rationale and the supporting data, and implementation remains a separately authorised, logged step.
What is the result of the comparison?
The result is not a general “best provider” but a documented decision for your defined case. It contains met and open knock-out criteria, the weights you chose yourself, verifiable answers, commercial scenarios and remaining risks. If two offers are still equivalent after this review, a limited pilot with acceptance and exit rules defined in advance yields more insight than additional decimal places in the score.
Which documents should you receive before signing the contract?
- a complete services and price formula with the behaviour when the budget changes;
- a role and access matrix including the payment profile;
- a conversion and data source definition;
- budget, approval and emergency rules;
- an example of decision and change documentation without another client’s data;
- reporting definitions;
- a list of external tools and dependencies;
- termination, data return and handover procedures.
A missing document is not an automatic exclusion. It does, however, remain visible as an open item until the provider clarifies it in a binding way.
Frequently asked questions
Is a Google Partner badge sufficient?
No. A badge proves a qualification or partnership. Also check the scope of services, control, evidence and the fit with your case.
How do I compare different pricing models?
Convert all models into annual costs under an identical scope of services and the same budget scenarios, and check how the fee behaves when the budget changes. maitiq can be put into the same calculation: the offer applies per account and does not change with your media budget.
Why do two offers quote different lead numbers?
Often because of the measurement definition: a primary conversion action appears in Conversions and can influence bidding when the campaign uses its standard goal; secondary actions normally appear only under All conversions, but if a campaign includes them through a custom goal, they can take effect regardless of their secondary status. The counting option “every” or “one” also changes the figure. The complete review chain is described in the conversion tracking audit.
Who should own the Google Ads account?
Keep administrator access on the client account and clarify the payment profile, linked services and external tools in the contract as well.
May the highest score win automatically?
No. Knock-out criteria that are not met remain grounds for exclusion or clarification, and open criteria are reported. The result supports a decision but does not replace taking responsibility.
Which weights are right?
Your business determines that on the basis of risk, organisation and goals. The matrix contains no preset weights, and there is no substantiated Swiss industry standard for them.
Sources and how to read them
This page draws on seven Google help pages and five commercial or directory pages. The Google help pages explain the platform’s functions and limits; the commercial and directory pages are publications by providers or publishers. Their rankings reflect those publishers’ choices and do not establish a neutral Swiss best-provider list, an endorsement of any provider, representative fee or risk norms, or rights to the account. All twelve sources are listed below. Information on how maitiq works is available at maitiq.com.
To compare fees and services, we additionally include your existing offer or your invoice; this information is not held in the Google Ads account.