PPC Analytics Pricing: The Audit Log Is on the Price List

PPC analytics pricing gets compared line by line — this plan is cheaper than that plan, this tier has more seats. That comparison misses the thing that actually decides your bill, and it misses it every time. Every platform in this market puts a meter on something. Six live price lists, read side by side, put six different things on the meter: a call, a credit, a client, a client’s ad spend, a data source, or nothing your team does all day.

That choice is worth more attention than the headline number, because it decides which behaviour gets more expensive as you grow. And in a regulated vertical, one of those behaviours is asking what happened.

Six price lists, six different meters

Every figure below was read off the vendor’s own live pricing page on 24 September 2026. No briefs, no summaries, no secondhand numbers.

Synter meters the call, and sells the record by the day

Synter’s pricing page states the model plainly: “Choose a subscription with included usage, seats, and workspaces for your team. Additional agent usage is metered at the listed rates.” Solo is $20 a month with $20 in claimable credits; Scale is $500 a month per workspace with $500; Custom is sales-led. Both self-serve plans include “1 advertiser account per platform” — worth reading twice if you run multi-client Meta, because that funnel points at Custom.

Then the catalog. One credit is $0.025, and the page lists what each action costs: reads (list, pull, BuiltWith) $0.05, light writes such as pause $0.05, an unpriced MCP, SDK, CLI or API call $0.25, campaign launch $25, image generation $2, video generation $27.50, pixel and CAPI ingest free.

Credit where it is due, because this part is genuinely good: Synter publishes a live rate map and lets you ask the price before you act. “You see what an action costs before it runs,” the page says, and forecast_tool_cost is the mechanism. It also commits to the thing every agency asks first: “Synter does not charge a percentage of your ad spend.” And “No fees. No retainers. Only pay for what you use” is an honest description of a per-call meter.

Here is where the meter lands on your compliance work. Scroll to the feature table’s governance section and the row reads:

That is a plan row. The window over which you can reconstruct what happened to a client’s account is a line in a pricing table, and the cheapest position of it is seven days. When a compliance officer asks in September who moved a budget in March, the answer on the entry tier is not “nobody” and not “here it is.” It is “out of window.”

Supermetrics meters AI credits. Its own FAQ says it doesn’t.

Supermetrics has spent fifteen years selling the most predictable licence in marketing analytics: you buy data sources and destinations, and the price does not move when your team does more work. For a reporting-heavy agency that is a real commitment, and the plans still deliver it — Starter from €49 a month, Growth from €199, the third tier from €499, Enterprise custom, with “3 data sources of your choice” and “3 accounts per data source” at entry.

But read the plan cards, not the FAQ, and a second meter appears. Every tier now carries a credit allowance: “Supermetrics Studio Includes 4,000 monthly AI credits”, then 12,000, then 18,000, then custom. And four sections further down the same page, under the question “Does Supermetrics charge based on usage?”, the company answers:

Both sentences are on one page, and both are the vendor’s. The honest resolution is not that anyone is lying: the licence is metered one way and Supermetrics Studio is metered another. The point is that the question an agency actually types — what will this cost me next month? — now has two answers on one page, and you have to know which meter applies to the work you are buying.

Adalysis meters the client’s ad spend. AgencyAnalytics meters the client.

Adalysis opens its pricing page with a promise of simplicity: “One platform for all your PPC work. One straightforward pricing plan.” The selector underneath it is not a seat count. It is “Maximum monthly ad spend (in USD)”, and the price climbs the ladder with it — $149 at the $50k band, $249 at $150k, $349 at $250k, $599 at $500k, $950 at $1m, $1,600 at $2m, $3,000 at $5m, custom above that.

What Adalysis gets right is worth saying out loud: every plan carries “Unlimited Google Ads accounts”, “Unlimited Microsoft Ads accounts” and “Unlimited users”. There is no per-account meter at all, which is exactly what a 40-account agency wants. And $149 for up to $50k a month of spend is genuinely cheap.

The mechanism is still worth naming: Adalysis’ revenue tracks the client’s media budget. The tool gets more expensive because the client spent more, not because your team did more. For an agency whose fee is a percentage of that same spend, that is a matched cost. For an agency on a flat retainer, it is a cost that moves on its own.

AgencyAnalytics chose the other variable. Its pricing page says: “One rate. Every feature. Only pay for the clients you actually have.” The rate is $20 USD per client per month billed annually, with no feature gating, unlimited branded reports, and as many of the 85+ integrations attached to each client as you need. We will say the obvious thing rather than let a reader say it for us: this is the same shape as Ott’s model. Meter the client, or the brand, and the evidence costs nothing extra at the margin.

TracerGram publishes almost our line. Then there is ours.

The cleanest statement of the argument in this post is not on Ott’s pricing page. It is on a competitor’s. TracerGram sells one plan at €250 a month with “One plan. Full access. No hidden fees” and “Unlimited events, no extra cost”. Six Telegram workspaces, no event meter, no credit catalog. A direct competitor of ours got there first and we are happy to quote it.

Our own page, for completeness, states it this way: “No charges per ad account, data source, or Telegram join. Limits are by brands and seats so teams know exactly what they get.” Starter is CHF 399 a month for up to 5 brands and 3 seats; Agency is CHF 749 for up to 20 brands and 10 seats; Agency Pro is priced on request. The only meter that moves is capacity.

Now the footnotes, because a pricing post that hides its own exceptions is not worth reading:

• The page carries a spend-scaling element. Its own words: “Pricing scales with your ad spend, ensuring you only pay for what you use. At 1% of ad spend above included amounts, the platform pays for itself.” That is not nothing, and anyone describing our model as flat with no spend component is describing a different page.
• We meter API requests. Agency includes “REST API & MCP access — 500,000 requests/month”; Agency Pro gets “generous usage (per agreement)”. So we are not the vendor with no meter at all. We are the vendor whose meter is not pointed at the record.
• Starter excludes things. Telegram conversion tracking and REST API & MCP access are Agency-tier items, not every-plan items. Add-ons run CHF 39 a month for an extra brand slot, CHF 30 a seat, CHF 50 an extra client viewer seat, CHF 199 an extra Telegram channel mapping.
• We do not publish a retention number. We have said that before and will say it here. The narrow claim in this post is smaller than that: the activity record is not a unit on our price list, so keeping it does not get more expensive when your team has a busy month. A discount that cannot be metered is a discount nobody can quietly reduce later.

What does one audit question cost?

Put the price lists aside and run the scenario your client’s compliance function actually runs. A campaign called Forex Signals Jan was paused on 14 March. The person who paused it has left. The client wants the before-and-after budget values and the name of whoever approved the change.

On a per-call meter, answering that costs something — one call, or a handful, at the published rate. It is small. As a monthly habit across forty client accounts and a regulator’s information request, it is a line item that only exists because someone asked a question. On a credit meter, the same. On a spend-band meter, the question is free but the seat at the table cost more because the client’s media budget grew.

And on the entry tier where the record window is seven days, the cost of the March question in September is not a fee. It is the question being unanswerable, and the unanswerable version is the expensive one, because the only substitute is a screenshot someone remembered to take.

The meter test: four questions for any price list

This works on any vendor, including us. It takes about five minutes and needs no sales call.

The price list is not the record

Everything above is about what a vendor charges for, and that is the whole of what a price list can tell you. It cannot tell you whether the export names a person or an account owner. It cannot tell you whether the client’s jurisdiction is attached to the event. It cannot tell you what the record looks like when the ad account itself gets restricted and the platform’s own history UI disappears.

Those answers live in the documentation, the settings page, and the export — the three surfaces a price list never shows you. Read the price list to learn what you will pay for. Then read the record to learn what you will actually have.

Two questions, in this order: what does this vendor put on the meter, and what does the meter refuse to measure? The second one is where the compliance work lives, and no pricing page has ever answered it.