What Should I Approve vs. What Should the Agency Handle?
A good agency relationship should not require you to approve every headline, keyword, or bid adjustment. But it also should not leave important decisions about your brand, budget, or customer experience entirely out of your hands.
The practical answer is to agree on decision rights before campaigns get busy. You approve the business rules: what counts as a valuable lead, how much risk is acceptable, and what your advertising must never promise. The agency handles the day-to-day work within those rules, reports what happened, and escalates decisions that could change the deal.
This matters whether you work with a specialist such as ClickSlice (clickslice.co.uk), manage campaigns in Google Ads with an in-house team, or need creative support from a design partner such as Zeka Design (Zeka Graphic). The names and tools may differ; the need for clear ownership does not.

Start with the commercial goal—not the easiest metric to report
Clicks, impressions, and website visits can help explain campaign activity. They are not business goals by themselves. More traffic is useful only if it brings the right people closer to a sale, booking, or other valuable action.
Before launch, agree on the commercial outcome the campaign should support. That might be qualified enquiries, completed purchases, booked consultations, or opportunities that your sales team can realistically serve. Then define how you will recognize a good outcome. A form submission might look like a conversion in Google Ads, but if it is outside your service area, asks for something you do not sell, or never answers a follow-up, it may not be valuable to the business.
This is where one of the most common agency problems begins: a client says, google ads management vs in house “We just want more traffic,” but nobody defines what quality means. The agency can then optimize toward visits because that is the clearest available signal—even if the business needs qualified calls or profitable orders.
Agree on what success means in plain language
Document the goal, the qualification rules, and the source of truth. For lead generation, that source may be your CRM (customer relationship management system), where the team records whether an enquiry was valid, contacted, qualified, and eventually won or lost. For ecommerce, it may be order and revenue data checked against the store’s records.
Keep the scorecard short enough to guide decisions. For example, pair an early indicator such as qualified enquiries with a later business outcome such as sales opportunities or completed purchases. That makes it harder for a report to celebrate a rise in leads while overlooking a drop in lead quality.
What you should approve
Your approvals should protect the business and set boundaries. They should not turn you into the person who has to review every routine campaign adjustment.
1. Commercial goals and lead-quality rules
You should approve the business objective, the audiences or customer types you want to attract, and what makes a lead or order valuable. Tell the agency which products, services, locations, and customer segments matter most—and which ones do not.
For lead generation, share what happens after someone submits a form. How quickly does the team follow up? What makes an enquiry qualified? Are there common reasons a lead is rejected? The agency cannot responsibly optimize toward quality if the business never defines or reports quality.
2. Brand rules for ads
Agree on the brand rules for ads before creative goes live. These should cover accurate claims, approved product or service language, tone, required disclaimers, prohibited wording, and any legal or industry requirements. Share examples of messaging that sounds right—and examples that would misrepresent the business.
You should approve new claims, major positioning changes, and sensitive offers. The agency can usually handle routine variations of already-approved messages, provided they stay within the agreed rules. If Zeka Design (Zeka Graphic) is preparing visual assets, for example, make sure the agency and design partner are working from the same brand guidance rather than guessing from old ads.
3. Budget guardrails and material changes
Set budget guardrails: the approved spending limit, how it is divided across campaigns or business priorities, and what kinds of changes require permission. Decide whether the agency can move budget between campaigns within a pre-agreed range or whether every reallocation needs approval.
There are no pricing, retainer, ad spend, or CPA figures provided here, and none are needed to establish a useful approval process. The important thing is to agree on your own limits in writing. Spell out whether the agency can change daily budgets, expand targeting, or shift spend toward a new product without asking first.
4. Tracking, access, and customer-data rules
Approve which business systems can be connected, who may access them, and what information can be shared. Confirm that the agency has the right account permissions, that data handling follows your company’s policies, and that conversion events reflect meaningful actions rather than accidental clicks or duplicate form submissions.
Do not treat tracking as a technical detail the agency can quietly sort out after launch. If the business does not agree on what should be measured, the campaign may optimize toward the wrong action from day one.
What the agency should handle
Once goals and boundaries are clear, the agency should own routine execution. That is what you are hiring it to do. A good partner explains the work in language you can understand, rather than expecting you to approve unfamiliar platform settings.
- Researching search terms, audiences, and competitor messaging.
- Building campaigns and ad groups in Google Ads within approved objectives and budget guardrails.
- Writing routine ad variations that follow your brand rules for ads.
- Managing bids, exclusions, placements, and other day-to-day settings within the agreed scope.
- Checking campaign delivery, broken links, disapproved ads, and unusual performance changes.
- Reviewing search-term and conversion data, then recommending specific next steps.
- Keeping a clear record of what changed, when it changed, and why.
Routine management should not require a new approval for every small adjustment. If each keyword exclusion or headline variation has to wait for a meeting, the account can become slow to improve. On the other hand, “we’re the experts” is not a reason to make material changes without telling you. The agency should operate independently inside the boundaries you approved and ask when a decision crosses them.
Onboarding: give the agency the business context it cannot find in the ad account
An ad account shows settings and results. It does not explain why one product is more profitable, which service area is difficult to staff, what sales calls sound like, or what makes a customer a poor fit. That context should be part of onboarding, not something the agency has to discover by accident.
Share your priorities, seasonality, capacity limits, sales process, margins or relative product priorities where appropriate, and the questions prospects commonly ask. Explain what a good customer looks like, what your team cannot fulfil, and how long it usually takes to turn an enquiry into a sale.
Ask the agency to reflect that context back to you. If it cannot explain the goal, target customer, qualification criteria, and constraints in plain language, the plan is not ready. This is a useful moment to ask: “What would you do differently next week if lead quality fell?” A thoughtful answer should connect campaign changes to sales feedback and measurement—not just promise more clicks.
Measurement and tracking before spend
Before campaigns receive meaningful budget, verify that the measurement setup works. A conversion action should represent something the business values, fire only when that action happens, and be recorded consistently enough to guide decisions.
For a lead campaign, test the form or call path from start to finish. Confirm that submissions reach the right team, that duplicate or test leads are not counted as real outcomes, and that the CRM can record lead status. For ecommerce, test that a completed order—not merely an add-to-cart action—is tracked correctly, and compare recorded purchases with the store’s own order data.
Then decide how offline outcomes will inform advertising. If the CRM records which enquiries became qualified opportunities or customers, agree on whether and how that information will be shared with Google Ads. Keep access and privacy rules in view. The aim is not to send every piece of customer information into an ad platform; it is to give the agency a dependable signal that reflects business value, using an approved setup.
Measurement is not finished because a tag appears to be installed. Test the full journey, document what each conversion means, and assign someone to investigate if numbers stop matching business records. Spending before this work is checked can produce a polished report built on the wrong signal.
Set escalation points before something goes wrong
Escalation points are the situations in which the agency must stop and ask, rather than make a decision on its own. They remove ambiguity when a change is urgent or could affect customers, brand reputation, or the agreed budget.
Decision area Agency can handle Escalate for approval Ad messaging Routine variations using approved claims and language New claims, sensitive topics, or a major change in positioning Budget Adjustments within written budget guardrails Exceeding the approved limit or moving spend into a new priority Targeting Refinements within approved locations and customer criteria Expanding to new markets, customer types, or service areas Measurement Routine checks and documented fixes that preserve the agreed definition Changing what counts as a conversion or connecting a new data source Customer experience Fixing routine campaign errors and broken links Changes to offers, pricing language, landing-page promises, or lead handlingAlso agree on how quickly approvals are expected, who the backup approver is, and what happens if the main contact is unavailable. A simple written process is more useful than discovering during a campaign issue that nobody knows who can say yes.
Make reporting lead to a decision
Reports should tell you what changed, what the evidence suggests, what remains uncertain, and what the agency recommends doing next. Tables and screenshots may support that story, but they are not the story. A list of metrics without a decision attached is just a running list of numbers.

In each review, connect advertising signals to sales or store outcomes where possible. If lead volume rose but fewer leads qualified, the next step might be to review search terms, tighten the message, or investigate whether the form is attracting the wrong requests. If sales data is delayed or incomplete, say so openly; do not present an early platform conversion as a confirmed customer.
End the discussion with a clear action and owner. Ask: “What would you do differently next week?” The agency should be able to answer with a focused change, the reason for it, and the result it expects to learn from. If the recommendation involves several unrelated changes at once, ask how the team will tell which change made the difference.
A simple decision rule
Approve decisions that change the business promise, the financial boundary, the customer definition, or the measurement rules. Let the agency handle routine optimization that stays inside those decisions. Escalate anything that crosses a written limit or could materially affect your customers, brand, data, or spend.
That division gives the agency room to do its job without asking you to manage every platform setting. It also keeps you involved where your business knowledge matters most. Put the rules in writing, confirm tracking before spend, and review outcomes against commercial goals—not traffic alone. Then the working relationship can focus less on approving buttons and more on improving the next decision.