Is Your Media Buyer Still Worth It in 2026?

Is Your Media Buyer Still Worth It in 2026?

Your media buyer is worth it only if they make the decisions the platform cannot. Those include the offer, the creative, the test plan, and what counts as a good customer. If most of their week goes to bids, budgets, and audience settings, you are paying for work the platforms now automate. Audit one week and find out.

THE PATTERN

What are you actually paying your media buyer to do?

Pay for decisions. Price campaign administration at what it now is: supervising the platform's automation. A busy change log can help explain activity, but it cannot establish value by itself. Ask one question: what got better this week because a person was involved?

In the engagements we run at Vista this year, we have watched operators without formal ad training build AI-run campaign systems. Those systems matched or exceeded what an experienced specialist believed was possible on the same account.

A pattern we see often is simpler: an owner pays a retainer or salary and cannot describe the buyer's weekly contribution. If you cannot name one decision your buyer made last week, the retainer is running on habit. Review the scope, even if the reports arrive on time and the relationship is friendly.

Agent-Does-the-Work. Vista uses this framework to separate execution that AI can perform from judgment a human must own. For paid advertising, review the automated delivery work first, then name the decisions your buyer remains responsible for.

We also see AI-run campaigns stall because the ad gives people no reason to act. Picture a contractor running "Free estimates, quality work" while every competitor says the same thing. No bid adjustment fixes that. The work that pays sits upstream: the customer, the promise, and what evidence would justify the next test.

Good buyers pass this review easily. A buyer who spots a weak promise and rebuilds the campaign around a better one earns the fee. A buyer whose explanation ends at settings and dashboard screenshots is billing you for the platform's work.

THE BASELINE

What does platform automation already cover?

Start with the automation actually enabled in your account, because available features and your chosen setup can differ. Then compare your buyer's work with that baseline. You need to know which tasks require manual attention, which tasks are automated, and which tasks require someone to approve the result.

Google describes Performance Max as using AI across bidding, budget optimization, audiences, creatives, and attribution. Meta's Advantage+ announcement says its most advanced optimizations turn on when advertisers use Advantage+ audience, placement, and campaign budget. That is the baseline for your audit.

The verdict follows: routine delivery work no longer justifies a retainer on its own. What the platforms cannot tell you is whether your offer is any good, or whether the leads turn into paying customers. That is where a human earns the fee.

What the buyer does	Who does it now?	What to pay a human for
Adjusts bids during routine delivery	Mostly the platform, if the automation is switched on	Selecting the business goal and handling justified exceptions
Moves budget between delivery opportunities	Mostly the platform, if the automation is switched on	Setting spend boundaries and deciding what warrants more investment
Rotates audiences or placements	Mostly the platform, if the automation is switched on	Defining eligible customers, exclusions, and customer signals from actual sales
Uploads variations of an existing ad	Mostly the platform, if the automation is switched on	Creating distinct concepts and approving the claims they make
Reports clicks and leads	Automated reporting; value still unproven	Connecting lead quality and completed sales to campaign decisions
Designs the offer and test plan	A human, always	Choosing a proposition, a test question, and a decision rule

Use the table as an audit map, with account settings beside it. Do not assume that every row is automated just because the campaign carries an AI label. A good buyer can show what is active, explain deliberate exceptions, and identify a business reason for work that remains manual.

This is the buyer side of the shift described in the AI pricing arbitrage window. The question for you is whether the service has evolved with the work. Faster execution can leave room for better judgment, but the agreement should say what that judgment includes.

THE HUMAN WORK

Which decisions should a good buyer still own?

Human judgment starts upstream. A good buyer should turn business knowledge into better offers, creative concepts, and test decisions. They should also make the limits of the evidence clear. That work remains valuable when delivery is automated, especially if you lack the time or experience to make those decisions yourself.

Start with the offer and the customer

Your buyer should understand why the intended customer would act now, and what might stop them. Ask them to explain the offer without opening the advertising account. They should be able to discuss the promise, the proof, the next step, and how each fits the business's ability to deliver.

For a service business, a relevant question might be whether the advertised work fits the jobs you can fulfill well. An attractive lead can still be a poor commercial fit. Give your buyer feedback from actual conversations, including objections and declined quotes, so their next recommendation reflects more than the ad report.

• Offer: What specific promise are we making, and can we deliver it?
• Creative: What distinct reason to act does this concept communicate?
• Measurement: What would count as a valuable customer outcome?
• Testing: What decision will this test help us make?

Require a test question before a new ad

Testing discipline means saying what you want to learn before spending money to learn it. Changing an image and a headline can be worthwhile, but the buyer should explain the question. Perhaps the test compares proof of workmanship with convenience, or an immediate booking with a request for more information.

Ask what stays consistent and what changes, then agree how the result will affect the next decision. A disciplined test can end with uncertainty, provided the buyer explains why. Keep a short record of the idea, the evidence, and the next action so the learning survives a change of staff or agency.

Connect reporting to what the owner knows

Your buyer should reconcile the campaign report with your experience of the customers it brings. When lead volume looks healthy but the team reports poor fit, investigate the discrepancy before approving another increase in spend. Ask what information is missing before accepting a recommendation to spend more. A chart cannot answer questions nobody has put into the review.

Buy a decision. A weekly report ends with a recommendation, its evidence, and its limits. A list of account changes is the receipt. The recommendation is what you paid for.

You also need someone willing to say the business is not ready to advertise more aggressively. If the offer is unclear or follow-up is unreliable, extra traffic may add work without resolving the problem. Our discussion of when paid advertising is premature gives that decision a place in the review.

THE AUDIT

What should you ask during a one-week audit?

One week will not tell you whether the campaigns work. It will tell you what your buyer actually does, what the automation handles, and where nobody is responsible. Ask for the evidence. Put those answers beside the scope you pay for so the review ends with a decision.

At the start of the week, request the current scope, the active automation settings, and a recent report. Add the latest creative brief and the record of completed tests. Read them alongside feedback from whoever handles inquiries or sales. That context keeps the review grounded in the work you are buying and the customers it brings.

1. What did you do last week that the account's enabled automation would not do? Ask for a decision and its business reason. A manual action can be valuable, but the explanation should connect it to an actual constraint, risk, or opportunity.
2. What did you change about our offer or our message, and why? Look for a distinct proposition or customer insight. An upload list shows production activity, while the reasoning should explain why the next ad should receive attention and spend.
3. Which recent test changed what we believe about our customer? Ask to see the original question and the conclusion. If the result was inconclusive, the buyer should say so and explain what information would help them make the next decision.
4. How do you check whether reported leads become customers who fit our business? Request the connection between campaign reporting and your team's experience. If that connection is missing, name who will supply the information and who will review it before recommending more spend.
5. What would make you pause, reduce, or redirect spending? Ask for conditions that you can both recognize. Include business capacity, poor lead fit, unreliable measurement, and an offer that needs revision. Agree who can act and who must be consulted.
6. What should we stop paying for, and where should that effort go instead? Invite the buyer to revise the scope around work that matters. A strong answer may remove routine administration and add creative development, customer research, or more careful test reviews.

Keep the review focused on evidence, even if an answer is disappointing. The buyer may need customer feedback or owner approval that your team has not supplied. Mark each gap as missing provider work, missing business input, or an unclear responsibility. That distinction helps you repair the agreement without blaming the wrong person.

At the end of the week, write a plain-language scope with named outputs and decision owners. Use which tasks to hand to AI to review execution separately from approval. The resulting agreement should make it easy to explain why you need the human involvement and how you will assess it.

THE VERDICT

When is the buyer worth keeping or replacing?

The scope must earn its fee. Keep a buyer who improves your decisions, learns from your customers, and can explain the boundaries of their responsibility. Revise or replace a service that cannot show those contributions. The fair standard applies to an agency, a freelancer, an employee, and an owner doing the work themselves.

What your audit reveals	Practical verdict	Next action
Clear judgment, decisive tests, and shared business evidence	Keep the buyer	Preserve those responsibilities in the scope
Strong expertise, but a scope dominated by routine work	Revise the agreement	Redirect effort toward offer, creative, and measurement
Activity reports without meaningful decisions or accountability	Replace the service	Ask once for a revised scope; if it is not credible, move on.
Missing owner inputs prevent progress	Repair your side first	Supply customer feedback, capacity limits, and approval ownership

Worth it if

The buyer improves the proposition, produces distinct creative ideas, and runs tests that lead to decisions. They check business feedback before recommending more spend and explain uncertainty. They also use automation sensibly, with a clear explanation of which settings they trust and which decisions they reserve for human review.

Not worth it if

The service mainly sells routine account activity and cannot explain what that activity contributes beyond enabled automation. Repeated reports arrive without a test conclusion, a creative recommendation, or a business decision. If the buyer cannot propose a credible revised scope, replace the service. Capable and responsive is not enough when the platform already does the job you are paying for.

Make any transition with named responsibility for account access, ongoing spend, and current tests. The owner still needs to know who is watching the account while the service changes. Replace unclear work with a clear operating arrangement, including who approves new creative and who decides when evidence warrants another investment.

Bring your audit to a Vista AI Lab session and work through which decisions you should own, automate, or keep buying. Bring the scope and recent test records so each decision starts from the work you actually receive each week.

Pay for the decision, not the dashboard.

COMMON QUESTIONS

Frequently asked questions

Can AI replace my media buyer completely?

Not completely. The platforms now automate much of the execution, but somebody still owns the offer, creative approval, spending limits, and business judgment. Review those responsibilities before changing the service. If you take them in-house, name a person with the time and information to do them.

Is an agency using automation a reason to cancel?

Using automation is not a reason to cancel an agency. Ask what the agency contributes around that automation, including offer development, creative concepts, measurement, and testing decisions. The important issue is whether its scope and fee reflect human judgment that your business still needs someone to own.

What should a weekly media buying report include?

A useful weekly media buying report includes the decisions made, the evidence behind them, and what happens next. It should connect campaign activity to customer quality and business outcomes where that information is available. It should also identify uncertainty, missing inputs, and the person responsible for resolving each open question.

Can a one-week audit tell me whether advertising works?

A one-week audit can reveal what your buyer does and how responsibilities are divided. It does not establish whether a campaign will deliver lasting results. Use the week to inspect the scope, recent test records, automation settings, and customer feedback, then agree how performance should be reviewed over time.

What if my buyer says creative is outside the scope?

If creative is outside your buyer's scope, decide who owns it and how their work connects to campaign learning. A delivery specialist can still be useful within a clear arrangement. Review whether the fee suits that narrower role, and make someone accountable for developing the offer and new creative concepts.

Should I manage advertising myself to save the fee?

Manage advertising yourself only if you can own the decisions the service currently handles, including review and follow-up. Compare the work required with the time you can realistically commit. Removing a fee saves money when responsibility remains clear and important judgment does not disappear during the change in ownership.