Agencies · Multi-Account Tools

AI Ad Optimization Tool for Agencies: How to Manage Many Client Accounts From One Dashboard

Start with the free native layer, buy only the cross-platform gap, and keep each client's rules, data and tax setup separate as your roster grows.

Ajay Kumar • •11 min read
Illustration of one control desk linked to many client ad account screens, with separate locked folders for each client's data

Quick answer: what should an agency look for in an AI ad optimization tool?

An agency AI ad optimization tool manages many clients' Meta, Google and other ad accounts from one place, with bulk launching, per-client rules, budget pacing and white-label reports. Set up Google manager accounts and Meta partner access first, since they are free. Then pay only for cross-platform rules, reporting and creative analysis, and check how pricing grows with combined client spend.

Start with the free layer: Google manager accounts and Meta partner access

Google Ads manager accounts and Meta business portfolios already give you multi-account control for free. A paid tool should fill the gap they leave, not repeat what they do.

Google Ads manager accounts (MCC)

A manager account lets you manage and view data for every Google Ads account linked to it. It can link up to 85,000 non-manager accounts in total, counting active, inactive and cancelled ones.

The active-account cap depends on the highest total monthly spend in the last 12 months. Google measures this at the top manager account, the one with no manager above it. Google sets the bands in US dollars, so the figures below are rough equivalents. Check Google's account limits page for exact figures.

Highest monthly spend (last 12 months)Active accounts allowed
Under about ₹9 lakh50
About ₹9 lakh to under ₹4.4 crore2,500
Above about ₹4.4 croreNo published cap

Two more limits matter for agencies. According to Google's manager account help page and third-party guides, a client account can be directly managed by up to 5 manager accounts, and the structure can go up to 6 levels deep.

Meta business portfolios with partner access

Meta now uses the term "business portfolio" for what it used to call Business Manager, and you manage it in Meta Business Suite. The client owns the portfolio and its ad accounts, then adds your agency's portfolio as a partner with full control or partial access. Partial access covers specific tasks such as managing ads. Meta Ads Manager also offers built-in automated rules. Check Meta's help centre for the current options.

What the native tools still leave out

For what the platforms' own AI handles inside each account, see what Advantage+ and Smart Bidding already do for free.

Set up access and offboarding before you connect any tool

Every AI tool connects through the access you already have, so fix access first. The safe model is simple. The client owns the portfolio, your agency gets partner access, and your staff are assigned inside your own portfolio.

Why partner access beats adding staff as people

Turn on two-factor authentication for everyone, as Meta advises. Only connect tools that use the official platform APIs. Never use shared logins.

Offboarding checklist when a client leaves

  1. Export reports, rule logs and change history from the tool.
  2. Pause or delete that client's rules so nothing acts on stale data.
  3. Disconnect the client's accounts from the tool so they stop counting towards spend-based billing.
  4. Remove your partner access in Meta and unlink the account from your MCC.
  5. Confirm the pixel, Conversions API (CAPI, server-side event tracking), catalogue and audiences sit in the client's own portfolio.
What breaks

Once partner access is removed, the tool loses its connection to that client's assets. Any rules, alerts and scheduled reports for that client stop working. Run the checklist before access is removed, not after.

Shortlist by agency type, not by "best overall"

The right tool depends on your channel mix and how many clients you run, not on a ranking. Most agency tool lists are written by vendors who rank themselves first.

Every agency tool should do 4 jobs well: bulk launching across accounts, per-client rules and pacing with alerts, cross-platform white-label reporting, and creative analysis that never mixes client data.

Agency typeMain gapTools to shortlist
Meta-heavy D2CBulk launch, creative analysis, profit-based rulesMeta-focused tools such as Madgicx and Bïrch (formerly Revealbot)
Google-heavy PPC or lead genAudits, bid and budget rules across MCCGoogle-focused tools such as Optmyzr
Multi-channelRules and reports across platformsBïrch, TheOptimizer (its website says it groups reports by time zone, currency or tags)
White-label resellerBranded reports at volumeReporting tools, Acquisio

Our rule of thumb for when to buy

  1. Native layer not set up yet: do that first. No tool fixes broken access.
  2. Combined client spend under about ₹10 lakh a month: native tools plus a reporting connector are usually enough.
  3. Fewer than 10 clients, all on 1 platform: native rules and MCC usually cover you.
  4. 10+ clients, 2+ platforms and ₹10 lakh+ combined spend: a tool is justified for the cross-platform gap.

Before signing, run any vendor through these 10 questions to ask before buying an AI ad tool.

Pricing maths: what the tool costs as your client roster grows

Spend-based pricing looks cheap at the start and grows with every client you add. Flat pricing costs more at the start and less as you grow.

Bïrch, for example, prices on the monthly ad spend of all connected accounts, not per account or per seat, according to several third-party reviews of its pricing page. The same reviews say tiers step up with spend, monthly plans add an overage fee above the tier, and automated rules sit on the Pro plan only. We could not confirm these details on bir.ch directly. Prices change, so check bir.ch/pricing before you budget. Other vendors publish flat monthly tiers, and some charge per seat.

Where spend-based and flat pricing cross over

Model the next 12 months of your roster, not today's. Entry prices on ranking pages hide this.

Worked example: one agency at three roster sizes

Step 1: Set two simple price models

Example: Tool A charges about 0.5% of combined connected spend. Tool B charges a flat ₹40,000 a month. These are round illustration numbers, not any vendor's price.

Step 2: Cost at each combined spend
Combined client spendTool ATool B
₹10 lakh₹5,000₹40,000
₹50 lakh₹25,000₹40,000
₹1.5 crore₹75,000₹40,000

Tool A always costs ₹500 per ₹1 lakh of spend. Tool B falls from ₹4,000 to ₹800 to about ₹267 per ₹1 lakh.

Cost per ₹1 lakh of spend
Tool B at ₹1.5 crore: ₹40,000 ÷ 150 lakh = about ₹267 per ₹1 lakh
Step 3: Find the crossover

Divide the flat fee by the spend-based rate. Above that combined spend, the flat tool is cheaper.

Crossover spend
₹40,000 ÷ 0.005 = ₹80 lakh a month
Step 4: Pass the cost on to clients

Split the bill by each client's share of spend. A client spending ₹30 lakh of ₹1.5 crore (20%) carries ₹15,000 of Tool A's ₹75,000. Show it as a line item in the retainer, so clients see it grow with their own spend.

Set rules per client and per country, never one template

A rule that protects one client's margin can push another client into losses. Thresholds must come from each client's break-even ROAS (return on ad spend), not from an agency-wide template.

Example: two clients both show a platform ROAS of 3.0. Client A sells prepaid in the UK and breaks even at 2.2. Client B sells in India with many COD (cash on delivery) orders refused at the door. After RTO and GST it breaks even at 3.3. A shared "scale above 2.5" rule grows A profitably and scales B into a loss.

What to set separately for each client

ScaleWin judges each account against its own break-even ROAS for this reason. For how pause, scale and fix decisions should follow from those numbers, see how an AI tool decides what to pause, scale or fix.

Tax and invoicing differ by each client's market

Platform tax follows the billing address on each ad account, so map it per client. Meta calls this the "Sold To" address. Who pays the platform decides whose tax ID matters.

MarketWhat Meta doesWhat to set up
IndiaApplies GST and TDS (tax deducted at source) where Facebook India sells and the Sold To address is in India. TDS does not appear on monthly-invoicing invoices.Client GSTIN on the account. Map TDS in your finance process.
AustraliaAdds GST for Australian Sold To addresses, whatever the registration statusInclude GST in media plans
CanadaAdds GST/HST when the business country is Canada. Quebec has separate QST rules.Check the province
EU and UKAdds VAT on business purchases when a local Meta office is the seller (outside Ireland). Usually recoverable.VAT number in account settings
US, Gulf, South-East AsiaVariesConfirm with a tax advisor

Google's tax help page says Google Ads adds taxes under local law and shows them in transaction history and on the monthly invoice. The table above summarises Meta's tax help pages at the time of writing. Tax rules change, so confirm each market on the pages below before you set up billing. Sources: Meta's tax help page and Google's tax help page.

What most guides get wrong about agency AI tools

Most ranking pages repeat old names, wrong claims and entry prices. Here is what to correct.

The data clause to add to client contracts

State whether the client's data may be used to improve models used for other clients. Say how history is exported when the client leaves and when the tool deletes it. Name the tools you connect. A client who later finds their data informing a competitor's campaigns has a fair complaint.

A 30-day rollout for a 10-50 client agency

Pilot on 3 clients, prove the numbers against real revenue, then expand. Connecting every client on day 1 multiplies any mistake across your roster.

  1. Week 1: Audit access. Every client owns their portfolio, your agency has partner access, 2FA is on and MCC links are clean.
  2. Week 1-2: Connect 3 pilot clients with different profiles, such as one COD-heavy, one prepaid and one lead-gen.
  3. Week 2: Set per-client rules from each break-even ROAS, in alert-only mode.
  4. Week 3: Compare tool-reported revenue with backend revenue after returns and RTO. Note the gap for each client.
  5. Week 4: Switch on actions for rules that held up, add the data clause to contracts and expand in batches of 5-10 clients.
Tip

Keep new rules in alert-only mode for 2 weeks. You see what the tool would have done without risking a client's budget.

For the wider picture of costs and features, see our complete guide to AI ad optimization tools.

Free tools for this topic

Key takeaways

  • Google Ads manager accounts and Meta partner access already give you free multi-account control and native rules. Pay only for the cross-platform gap.
  • Many tools price on the combined spend of every connected client, so your bill rises with your roster. Model it at next year's spend, not today's.
  • Clients should own their business portfolios and give your agency partner access. Never use shared passwords.
  • Set rules per client and per country. A COD-heavy client in India or the Gulf needs different thresholds from a prepaid UK or US client.
  • Add a data clause to client contracts before using any tool that learns across accounts.

Ai ad optimization tool: frequently asked questions

How many client accounts can a Google Ads manager account (MCC) hold?

Up to 85,000 linked accounts in total, including inactive and cancelled ones. The active cap depends on the highest monthly spend in the last 12 months at the top manager: 50, 2,500, or no published cap at the highest band. Google sets the bands in US dollars, so check its help page.

Should my agency own the client's Meta ad account or use partner access?

Use partner access. The client should own the business portfolio and ad accounts and add your agency portfolio as a partner. This keeps ownership clean and makes offboarding simple for both sides.

How do AI ad tools charge agencies?

There are three common models: a share of combined connected ad spend, flat monthly tiers, or per seat. Spend-based pricing rises with your roster, so model the cost at next year's combined client spend before you sign.

Is Revealbot the same as Bïrch?

Yes. Revealbot now operates as Bïrch, and revealbot.com redirects to bir.ch. Third-party reviews date the rebrand to October 2024 and say pricing is based on the combined spend of all connected accounts. Check bir.ch/pricing to confirm.

What happens to my automations if a client removes our partner access?

The tool loses access to that client's assets, so rules, alerts and scheduled reports stop working. Export history and disconnect the client in the tool before access is removed.

Do I need an AI tool if I already use Google MCC and Meta Business Suite?

Not always. Below about ₹10 lakh of combined spend, or with fewer than 10 clients on one platform, native tools are often enough. A tool earns its cost when you need rules, reports and creative analysis across several platforms.

Sources

Ajay Kumar
Founder, ScaleWin · 12+ years in performance marketing

Ajay is the founder of ScaleWin. He works hands-on with Meta Ads, Google Ads and SEO for Indian D2C and service brands, and built ScaleWin so that the daily work of a performance marketer - reading numbers, pausing losers, scaling winners - happens automatically.

Drafted with AI assistance. How these articles are written

Let ScaleWin read your ad data every day

ScaleWin checks your Meta and Google Ads daily, judges every ad on profit and tells you exactly what to change.

Start free, no card

Get marketing tips that grow your business

One practical email a week on ads, SEO and growth. No spam, no fluff. Unsubscribe anytime.