---
title: "The 1am Spreadsheet Is Coming Back for AEO Agencies"
description: "I know what happens when agencies sell judgment but deliver it through manual account work. AEO across a dozen domains is heading for the same 1am spreadsheet."
url: "https://groas.com/post/i-managed-a-dozen-clients-ads-by-spreads"
image: "https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/fdea73fc-3d9d-492b-ab11-e5ad2819188a.png"
published: "2026-10-08T05:43:17.822Z"
modified: "2026-10-08T05:43:17.964Z"
---

October 8, 2026 · 9 min read

# The 1am Spreadsheet Is Coming Back for AEO Agencies

[Alexander PerelmanHead Of Product @ groas](https://groas.com/author/alexander-perelman)

![A tired man at a laptop at 1am as dozens of colour-coded spreadsheet tabs spill out of the screen and creep across his apartment.](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/fdea73fc-3d9d-492b-ab11-e5ad2819188a.png)

In this article

1. [The master list looked like leverage](#the-master-list-looked-like-leverage)
2. [One changed offer, another rebuild](#one-changed-offer-another-rebuild)
3. [The retainer could not cover the hours](#the-retainer-could-not-cover-the-hours)
4. [What I would not sell by the hour again](#what-i-would-not-sell-by-the-hour-again)

At 1:14 a.m., in a composite of my spreadsheet years, twelve client tabs are open and the laptop fan is running. Another search term waits for a decision. I copy junk into a master negative list, one tab at a time. Dog food brand, plumber, another plumber, SaaS startup. Same interface, different logos. The apartment is quiet. This feels like the job.

Agencies selling AEO across ten or more client domains are about to recognize that night. Swap search terms for buyer prompts. Swap negative lists for schema and content fixes. Add a client CMS login for every domain. I know why the model is appealing: the first account rewards care, and the second looks enough like it that you can reuse your work. Then the tabs multiply. **The work does not get lighter with practice. It gets heavier with every domain you add.**

## The master list looked like leverage

I had a system. Every client got a tab. Every tab had the same columns: raw term, added as negative, added to master. The masters were the prized possession. One for ecommerce, one for home services, one for the SaaS thing nobody could describe in under forty words. The doctrine, reflected in [advice to keep a master negative list per industry](https://searchengineland.com/negative-keywords-ppc-449123), was that this was leverage. Build the list once, reuse it for the next client in the vertical, charge for expertise you had already banked.

It worked until I counted the hours. Say twelve accounts produce 300 to 400 new search terms between them in a week. At two minutes a term to read, decide and file, that is roughly ten to thirteen hours of filing before anyone changes a bid or writes an ad. A master list shortens some decisions. It does not make the new terms read themselves, or tell you whether the same word means the same thing to two different clients.

I told myself that distinction was craft. I knew which plumber needed _jobs_ blocked and which one needed _hiring_ blocked. That knowledge mattered. The tenth time I blocked _free_ for a dog food brand, though, I was not making a fresh strategic call. I was doing data entry with opinions. **Clients paid for the judgment. The copying ate the night.**

Then came the morning a home services account stopped behaving like the account I had built. I checked the bids, the keywords, the search terms. Only after working through what I controlled did I open the site. The client had swapped the landing page without telling me. The page I had mapped the ads to was gone. Now I had to find out what had changed, what no longer matched the ads and which parts of the account needed rebuilding.

A landing page change can damage ad relevance and the experience behind a click, both of which matter to how an ad competes. I used to tell clients Quality Score was a report card. I was wrong to treat it as something you could glance at after the fact. The important bit was the mismatch: the ads still promised one thing, and the page now presented another. I had built a neat account around a moving target and learned about the move last.

**A silent site edit is not a small maintenance task when your delivery depends on the old page.** That morning is the part I would carry into any AEO contract.

## One changed offer, another rebuild

The SaaS startup made the same lesson harder to ignore. Its offer changed every quarter. Each change sent me back through the account: ad groups, match structures, negatives, landing page mapping. I kept the old learnings where I could and crossed out what no longer applied. Then I billed rebuild hours and told myself that was strategy.

Some of it was. Deciding what the new offer meant for targeting required a human conversation. After that came the long, unglamorous transfer of the decision into every place the old one lived. **The judgment happened once. The implementation kept reappearing.** A spreadsheet does not know that a client has changed its mind; it faithfully preserves yesterday’s answer until someone opens it.

Now put that offer change inside an AEO engagement. The remap is not just keywords. It is the buyer prompts you track, the pages meant to answer them, the structure that helps describe those pages, and the evidence behind what they say. A client changes its positioning and the prompt map may still run. Its report may still arrive on time. That is the danger: a workflow can look wonderfully orderly while measuring an offer the client no longer sells. Someone has to connect the new positioning to the old prompts, then follow that change through the pages and the reporting. Otherwise each step can be completed on schedule while the work as a whole stays pointed at yesterday’s buyer.

I keep seeing the reassurance I used to give myself, rewritten as [advice for delivering AEO services](https://agency.rankplusplus.com/guides/how-to-offer-aeo-services-to-your-clients-complete-guide): segment your clients, reuse templates, manage the work in Asana, automate the checks. I understand the attraction. My master negative lists were templates. They saved me from starting at zero. They did not tell me when the buyer language underneath them had shifted.

A practitioner reviewing AEO platforms made the other half of the problem plain: after a tool finds gaps, [someone still has to prioritize them, gather evidence, create the page, get approval and publish it](https://www.linkedin.com/pulse/aeo-tool-market-one-heres-how-choose-right-platform-anita-moorthy-vmlwf). Track the wrong prompts and a tidy report can give you confidence in the wrong work. I knew that feeling from PPC. The columns were filled in; the client’s offer had moved on. **A completed checklist is not proof that you worked on the right problem.**

## The retainer could not cover the hours

The spreadsheet looked less romantic when I put a price beside it. Take a home services client paying $1,500 a month, within the [flat-retainer range of $500 to $5,000](https://www.commercepundit.com/blog/google-ads-pricing-budgeting-tips/). If the account needs six hours a week for search terms, bids, ad tests and landing page checks, that is about twenty-four hours a month. At a $75-an-hour cost basis, the work costs $1,800 before I answer an email. The retainer has already lost.

There are only so many ways to live with that arithmetic. Charge more, spend less time or find a way to perform the repeated work without a person opening every account. We often chose the second option while describing it as a process. Check in every two weeks. Let the small things wait. Call the meeting an optimization review. I am not proud of the euphemism, but I recognize the calendar it came from.

AEO can make the same gap look like margin. [Monitoring and reporting packages are priced at $500 to $800 a month, while packages adding content, technical fixes and schema run $1,500 to $3,000](https://www.airanklab.com/blog/best-aeo-trackers-agencies-marketing-teams). The tracker may cost $79 to $250 a month per brand. Those figures can leave room for a business. They do not tell you how many approvals, page edits or CMS handoffs the month will contain. Nor does a cheaper tracker shorten an approval chain. If the fix is ready but nobody can publish it, the account still needs attention, and that attention still has to come from somewhere.

Give twelve domains four hours each for prompt review, edits, publishing and approvals. That is 48 hours a month. If those hours could otherwise be sold at the [$100 to $250 hourly rates quoted for AEO work](https://higoodie.com/blog/aeo-agencies/), the opportunity cost is hard to ignore. More important, four hours is a ceiling you invented for the quote, not a promise the client’s site has made to you. A template change does not check your calendar before stripping a page of the structure you were relying on.

Percentage-of-spend fees could hide this problem in PPC. Spend went up, the fee went up, and the number of hours might not. What the deck called alignment, I called a raise for doing nothing. AEO does not even give that model a spend figure to take a cut of. It gives you domains, each with its own offer, pages, approvals and surprises. **If the price is fixed while manual hours grow with every domain, the twelfth client exposes the margin you thought you had on the first eleven.**

## What I would not sell by the hour again

Eventually I stopped thinking of those late-night tabs as evidence of dedication. The question was which decisions genuinely needed me and which ones only needed to happen reliably. Obvious junk terms did not become more valuable because I copied them myself. Borderline terms did need a judgment call. Bid and budget changes needed guardrails. A changed offer needed a conversation, because no rule could decide whether a client now wanted demo requests instead of free-trial signups.

That is the split I would want in an AEO operation. I would not quote by the hour for someone to re-run buyer prompts across clients, hand-check the same technical structure each week, move approved copy into a succession of CMS logins and write a report explaining the changes. Those are not four independent little chores. They are the same recurring obligation seen from four tabs. Leave any one of them to a fortnightly check-in and the rest of the process can keep reporting on an out-of-date page.

I would keep people close to the calls the software cannot make alone: what the client now sells, which buyer question deserves an answer, whether a proposed page has evidence worth publishing, and who has authority to approve it. Then I would expect an execution system to carry the repeated work between those calls. An approval can wait on a person without the whole operation forgetting the page exists. I would want to see that distinction clearly: what needs a decision, what is waiting for permission, and what should already be live. **Buying another hour of attention is not the same as building a way to notice change.** I learned that distinction much later than I should have.

One PPC shop [describes moving repetitive execution onto an engine while scaling to 40 clients without hiring new account managers](https://groas.com/post/ppc-agency-scaled-40-clients-without-hiring-google-ads-execution-engine). The humans stayed client-facing. The nightly work did not need a person per account. That is the split behind [my argument against the manual, percentage-of-spend agency model](https://groas.com/post/google-ads-agency-unnecessary-2026-automation-vs-agency). An agency adding AEO should buy execution software rather than hire its way into another row of tabs. **If the delivery model needs a person per domain for repetitive work, it is rebuilding my 1am spreadsheet with someone else’s logo on it.**

I will not hand you a vendor shortlist. I will tell you what I would look for before trusting software with those domains. It would need to run the buyer prompts without someone opening each client account, show me when a page stopped appearing, notice a site change that makes yesterday’s map unreliable, and carry an approved fix through to publication with a readable record of what changed. If a person still has to copy, paste, publish and report for every domain, the software has given them a prettier tab. That may make the night more comfortable. It does not make it shorter.

I can still picture the old spreadsheet: twelve tabs, color-coded, a cursor blinking in cell A1 while the laptop fan runs. One more term waits to be filed.

## Frequently Asked Questions

### Does running AEO for a dozen client domains get easier with practice?

No. Unlike work where repetition builds efficiency, AEO delivery across many domains adds a new set of buyer prompts, schema and content fixes, and a separate CMS login for every domain. The work gets heavier with every domain you add, not lighter.

### Are master negative keyword lists a good way to save time across PPC clients?

A master negative list shortens some decisions, but it does not make new search terms read themselves or tell you whether the same word means different things to two clients. Twelve accounts can produce 300 to 400 new terms a week, which is roughly ten to thirteen hours of filing before any bid or ad changes.

### What happens when a client changes their landing page without telling the agency?

The ads can end up promising one thing while the new page presents another, which damages ad relevance and the experience behind a click. Because the agency built the account around a moving target and learned about the change last, it has to find out what changed, what no longer matches the ads and which parts of the account need rebuilding.

### Why is a client's offer change a problem for AEO workflows?

When positioning changes, someone has to connect the new offer to the old buyer prompts, then carry that change through the pages and the reporting. Otherwise the prompt map and reports can keep arriving on schedule while measuring an offer the client no longer sells, so a completed checklist is not proof of work on the right problem.

### How can a flat AEO or PPC retainer become unprofitable?

If a client pays a $1,500 monthly retainer but the account needs about six hours a week of manual work, that is roughly twenty-four hours a month, which costs $1,800 at a $75-an-hour cost basis. The agency is then tempted to cut service, for example by checking in only every two weeks, while calling that a process.

### Why doesn't a percentage-of-spend pricing model work for AEO?

Percentage-of-spend fees rise automatically when spend rises, but AEO has no spend figure to take a cut of. Instead it gives you domains, each with its own offer, pages, approvals and surprises, so if the price is fixed while manual hours grow with every domain, the twelfth client exposes the margin you thought you had on the first eleven.

### Which AEO tasks should agencies automate and which should stay with people?

Software should handle the recurring execution: re-running buyer prompts, hand-checking technical structure, moving approved copy through CMS logins and writing reports. People should stay close to the calls software cannot make alone, such as what the client now sells, which buyer question deserves an answer, whether a page has evidence worth publishing and who has authority to approve it.

### What should an AEO agency look for in execution software?

It should run buyer prompts without someone opening each client account, show when a page stopped appearing, notice a site change that makes the old prompt map unreliable, and carry an approved fix through to publication with a readable record of what changed. If a person still has to copy, paste, publish and report for every domain, the software is just a prettier tab.

## Pay For Results, Not For Hours

Businesses buy the outcome, agencies resell it, and groas answers for it either way.

[See If You Qualify](https://groas.typeform.com/to/xC1bQNUT)

## Structured data

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://graph.groas.com/entity/groas","name":"groas","url":"https://groas.com/","logo":{"@type":"ImageObject","url":"https://groas.com/icon-512.png","width":512,"height":512},"sameAs":["https://www.linkedin.com/company/groas/"]},{"@type":"WebSite","@id":"https://groas.com/#website","name":"groas","url":"https://groas.com/","inLanguage":"en-US","publisher":{"@id":"https://graph.groas.com/entity/groas"}},{"@type":"WebPage","@id":"https://groas.com/post/i-managed-a-dozen-clients-ads-by-spreads#webpage","url":"https://groas.com/post/i-managed-a-dozen-clients-ads-by-spreads","name":"The 1am Spreadsheet Is Coming Back for AEO Agencies","description":"I know what happens when agencies sell judgment but deliver it through manual account work. AEO across a dozen domains is heading for the same 1am spreadsheet.","inLanguage":"en-US","isPartOf":{"@id":"https://groas.com/#website"},"about":{"@id":"https://graph.groas.com/entity/groas"},"primaryImageOfPage":{"@type":"ImageObject","url":"https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/fdea73fc-3d9d-492b-ab11-e5ad2819188a.png"},"breadcrumb":{"@id":"https://groas.com/post/i-managed-a-dozen-clients-ads-by-spreads#breadcrumb"},"dateModified":"2026-10-08T05:43:17.964Z","mainEntity":{"@id":"https://groas.com/post/i-managed-a-dozen-clients-ads-by-spreads#article"}},{"@type":"BreadcrumbList","@id":"https://groas.com/post/i-managed-a-dozen-clients-ads-by-spreads#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://groas.com/"},{"@type":"ListItem","position":2,"name":"Blog","item":"https://groas.com/blog"},{"@type":"ListItem","position":3,"name":"The 1am Spreadsheet Is Coming Back for AEO Agencies","item":"https://groas.com/post/i-managed-a-dozen-clients-ads-by-spreads"}]},{"@type":"BlogPosting","@id":"https://groas.com/post/i-managed-a-dozen-clients-ads-by-spreads#article","url":"https://groas.com/post/i-managed-a-dozen-clients-ads-by-spreads","mainEntityOfPage":{"@id":"https://groas.com/post/i-managed-a-dozen-clients-ads-by-spreads#webpage"},"isPartOf":{"@id":"https://groas.com/blog#blog"},"headline":"The 1am Spreadsheet Is Coming Back for AEO Agencies","description":"I know what happens when agencies sell judgment but deliver it through manual account work. AEO across a dozen domains is heading for the same 1am spreadsheet.","datePublished":"2026-10-08T05:43:17.822Z","dateModified":"2026-10-08T05:43:17.964Z","image":{"@type":"ImageObject","url":"https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/fdea73fc-3d9d-492b-ab11-e5ad2819188a.png"},"author":{"@id":"https://groas.com/author/alexander-perelman#person"},"publisher":{"@id":"https://graph.groas.com/entity/groas"},"keywords":"aeo, spreadsheet, account, 1am spreadsheet, work, client, page, hours","wordCount":1901,"inLanguage":"en-US"},{"@type":"Person","@id":"https://groas.com/author/alexander-perelman#person","name":"Alexander Perelman","jobTitle":"Head Of Product @ groas","description":"Ex Goldman Sachs and Ex Stanford Computer Science","image":"https://groas.com/media/blog/4f17cac3a81acc224cc1b2cfab8f94bbc295a3089d13e8421125771ef8d4064f.jpg","worksFor":{"@id":"https://graph.groas.com/entity/groas"},"sameAs":["https://www.linkedin.com/in/alexander-433793253/"]},{"@type":"FAQPage","@id":"https://groas.com/post/i-managed-a-dozen-clients-ads-by-spreads#faq","mainEntity":[{"@type":"Question","name":"Does running AEO for a dozen client domains get easier with practice?","acceptedAnswer":{"@type":"Answer","text":"No. Unlike work where repetition builds efficiency, AEO delivery across many domains adds a new set of buyer prompts, schema and content fixes, and a separate CMS login for every domain. The work gets heavier with every domain you add, not lighter."}},{"@type":"Question","name":"Are master negative keyword lists a good way to save time across PPC clients?","acceptedAnswer":{"@type":"Answer","text":"A master negative list shortens some decisions, but it does not make new search terms read themselves or tell you whether the same word means different things to two clients. Twelve accounts can produce 300 to 400 new terms a week, which is roughly ten to thirteen hours of filing before any bid or ad changes."}},{"@type":"Question","name":"What happens when a client changes their landing page without telling the agency?","acceptedAnswer":{"@type":"Answer","text":"The ads can end up promising one thing while the new page presents another, which damages ad relevance and the experience behind a click. Because the agency built the account around a moving target and learned about the change last, it has to find out what changed, what no longer matches the ads and which parts of the account need rebuilding."}},{"@type":"Question","name":"Why is a client's offer change a problem for AEO workflows?","acceptedAnswer":{"@type":"Answer","text":"When positioning changes, someone has to connect the new offer to the old buyer prompts, then carry that change through the pages and the reporting. Otherwise the prompt map and reports can keep arriving on schedule while measuring an offer the client no longer sells, so a completed checklist is not proof of work on the right problem."}},{"@type":"Question","name":"How can a flat AEO or PPC retainer become unprofitable?","acceptedAnswer":{"@type":"Answer","text":"If a client pays a $1,500 monthly retainer but the account needs about six hours a week of manual work, that is roughly twenty-four hours a month, which costs $1,800 at a $75-an-hour cost basis. The agency is then tempted to cut service, for example by checking in only every two weeks, while calling that a process."}},{"@type":"Question","name":"Why doesn't a percentage-of-spend pricing model work for AEO?","acceptedAnswer":{"@type":"Answer","text":"Percentage-of-spend fees rise automatically when spend rises, but AEO has no spend figure to take a cut of. Instead it gives you domains, each with its own offer, pages, approvals and surprises, so if the price is fixed while manual hours grow with every domain, the twelfth client exposes the margin you thought you had on the first eleven."}},{"@type":"Question","name":"Which AEO tasks should agencies automate and which should stay with people?","acceptedAnswer":{"@type":"Answer","text":"Software should handle the recurring execution: re-running buyer prompts, hand-checking technical structure, moving approved copy through CMS logins and writing reports. People should stay close to the calls software cannot make alone, such as what the client now sells, which buyer question deserves an answer, whether a page has evidence worth publishing and who has authority to approve it."}},{"@type":"Question","name":"What should an AEO agency look for in execution software?","acceptedAnswer":{"@type":"Answer","text":"It should run buyer prompts without someone opening each client account, show when a page stopped appearing, notice a site change that makes the old prompt map unreliable, and carry an approved fix through to publication with a readable record of what changed. If a person still has to copy, paste, publish and report for every domain, the software is just a prettier tab."}}]}]}
```