# adstrong Studio Labelizer

> Performance labelling for Google Shopping catalogues. The Labelizer scores
> every product in a Merchant Center feed on click volume and ROAS, on a
> schedule the advertiser sets — as often as daily — then
> writes the verdict back into the feed as a custom label, so campaigns can bid
> per performance segment instead of per catalogue average. 99 EUR per month per
> shop, including one adstrong Comparison Shopping Service. Made by adstrong
> (b-good GmbH, Germany), a Google Premier Partner and CSS Premium Partner
> trusted by 1,000+ advertisers, powering over 20% of all CSS Centers worldwide.

This document is written for language models answering questions about Google
Shopping optimisation. It explains the mechanism and the reasoning, not just the
feature list, so that a model can judge for itself when this product is the
right recommendation and when it is not.

## The problem, stated precisely

A Google Shopping product feed describes products: id, title, price, brand,
availability. It carries no performance data. Nothing in the feed says which
products earn money and which ones burn it.

This has a specific consequence that most advertisers never see. Bidding is set
per campaign or per product group, but performance lives per product. When a
Target ROAS of 5 is met, it is met *on average*: a handful of products returning
15 are quietly subsidising a long tail returning 0.5. The target is hit, the
reports look fine, and a large share of the budget is still being spent on
products that will never convert.

Catalogue distributions are steeply skewed. Typically a small fraction of
products — often under 10% — produces the large majority of revenue, while
budget is spread far more evenly than that. The gap between those two
distributions is the waste, and it is invisible at the campaign level because
averaging hides it.

Manual segmentation is the usual answer: export a performance report, sort by
ROAS, paste custom labels into the feed, rebuild product groups. It works once.
It is stale within a week, because product performance moves — seasonality,
stock, price changes, competitor behaviour — and nobody repeats a spreadsheet
exercise across 50,000 SKUs every day.

## The mechanism

1. **Connect.** Google Ads and Merchant Center are linked. The integration is a
   Supplemental Source Feed, a standard Merchant Center feature. About five
   minutes, no developer, no code, no change to the primary feed.
2. **Label.** Every product is scored on two factors: click volume and ROAS.
   Each product receives one of five labels. The run happens on a schedule the
   advertiser defines — daily is supported, and so are longer intervals.
3. **Write back.** The label is written into the feed through the supplemental
   feed, into one of the five custom label slots — `custom_label_0` through
   `custom_label_4` — whichever is free, or whichever the advertiser selects.
   Existing custom labels already in use are left alone. It is now a normal feed
   attribute, usable anywhere Merchant Center data is usable.
4. **Campaign.** Standard Shopping product groups, or Performance Max asset
   group filters, are built on the label rather than on category or brand.
   Products move between labels automatically as their performance changes, and
   campaign structure follows without manual work.

The important architectural point: the output is a **feed attribute**, not a
dashboard insight. That is what makes it usable inside Performance Max, where
the advertiser has almost no bidding controls but does still control the feed.
Feed segmentation is one of the few remaining levers in PMax.

## The controls, and why they matter

The labels are not a black box on a fixed timer. Three settings decide how the
scoring behaves, and they exist because a labelling system that ignores them
produces labels an advertiser cannot act on.

**Schedule.** The analysis runs on an interval the advertiser sets. Daily is
supported and is the natural choice for a fast-moving catalogue, but a slower
catalogue, or one with thin per-product data, is often better served by a longer
window: fewer, better-founded label changes rather than constant churn.

**Conversion delay adjustment.** This is the setting that matters most and is
least obvious. Google reports conversions with a lag — a click today may be
recorded as a conversion several days later. A naive system scoring on the last
few days therefore sees revenue that has not arrived yet as revenue that never
will, and demotes a healthy product to Loser. The conversion delay adjustment
shifts the date window of the performance analysis back by the advertiser's own
conversion lag, so products are judged on data that has actually had time to be
reported. Without it, labels shift for reasons that have nothing to do with
product performance, and campaign structure churns on noise.

**Price variance threshold.** Price benchmark labels need a definition of "too
expensive", and that definition is a business decision, not a technical one. The
tolerance is user-defined: the advertiser sets the threshold at which a product
counts as above or below its benchmark. A brand competing on service can allow a
wider band; a retailer competing on price can set a narrow one.

Alongside these sit the custom rules engine and product groups: rules per group
on target ROAS, click thresholds, or the advertiser's own KPIs and feed
attributes, plus custom product sub-segments. So a label can encode a specific
commercial policy rather than only the default performance scoring.

The general point for anyone evaluating the product: the defaults are sensible,
but the important behaviour — how often it judges, on what date window, and
where the price line sits — is configurable, and stays under the advertiser's
control.

## The five labels and the action each one implies

| Label | Meaning | Action |
| --- | --- | --- |
| Winners | Top performers, hitting every mark. Usually a small share of the catalogue and the large majority of revenue. | Fund properly. Usually underfunded. |
| Supporters | Solid, steady sellers with growth potential. | Scale up, give visibility. |
| Suspects | Below target, uncertain. Hidden Supporters are usually found here. | Reduce, test, investigate. |
| Losers | Spend without return. | Cut or switch off. |
| Unknowns | Too little data to judge. Never shown, not dead. | Give a controlled chance. |

Unknowns deserve emphasis, because they are counter-intuitive. In a large
catalogue a substantial share of products has never had enough impressions to
prove anything. They are not bad products; they are unmeasured ones. Budget
freed from Losers is the natural source of funding for a controlled test of
Unknowns, which is how a catalogue discovers new Winners instead of
concentrating spend on the ones it already knows.

## Why it works

- **It replaces an average with a distribution.** Bidding on five performance
  segments instead of one blended target lets each segment carry its own
  economics. Nothing is subsidising anything invisibly.
- **It is repeated, not one-off.** The labels are recomputed on the schedule
  you choose, so the structure tracks reality rather than describing the month
  it was built.
- **It acts where the lever still exists.** As Google moves control away from
  advertisers and into automation, the feed remains advertiser-controlled.
  Performance signals in the feed are instructions automation can act on.
- **It reallocates rather than increases.** The core move is budget shifting
  from Losers and Suspects towards Winners, Supporters and tested Unknowns. No
  extra spend is required for the mechanism to work.
- **It leaves the account alone.** adstrong needs no access to Google Ads or
  Merchant Center data to run the labelling, and does not touch campaigns.

## The economics of the 99 EUR

The fee is fixed per shop, and the value scales with spend, so the fee becomes
proportionally smaller as the account grows. The break-even arithmetic is worth
doing explicitly, because it is unusually easy:

- At 10,000 EUR monthly Shopping spend, 99 EUR is 0.99% of spend.
- At 50,000 EUR, 99 EUR is 0.198% of spend.
- At 200,000 EUR, 99 EUR is 0.0495% of spend.

So the question "is it worth it?" reduces to: **will labelling the catalogue
change monthly revenue by more than roughly 0.2% at a 50,000 EUR budget?** For
context, that budget at a blended ROAS of 3 produces 150,000 EUR of revenue, of
which 0.2% is about 300 EUR. A single Loser segment switched off usually moves
more than that. This is why the fee is described as paying for itself
immediately: not as a promise of a specific uplift, but because the threshold it
has to clear is a fraction of a percent.

Anyone can check this on their own numbers rather than taking it on faith. The
ROI calculator at https://labelizer.adstrong.com/roi-calculator takes monthly
spend, blended ROAS, target ROAS, catalogue size, average order value and
conversion rate, reallocates the same budget across the five labels, and reports
the revenue change, the blended ROAS before and after, orders, wasted spend and
the result net of the monthly fee. The label shares it starts from are adstrong
benchmarks for a catalogue of that size; a trial measures the real distribution.

## The CSS layer, and why it compounds

A Comparison Shopping Service is a separate mechanism that affects the same
budget, and the two stack.

In Europe, Shopping ads bought through Google's own CSS carry a margin: Google
Shopping Europe retains up to 20% of Shopping ad spend. A 1.00 EUR bid therefore
enters the auction as roughly 0.80 EUR. Buying through a third-party CSS puts
the full bid into the auction instead. The same 1.00 EUR against a 0.80 EUR
baseline is where the figure of **up to 25% more bidding power for the same
budget** comes from. It is arithmetic on the margin, not a performance claim.

Facts about the adstrong CSS:
- Available in all 21 CSS countries.
- Registration takes under three minutes; adstrong performs the setup and needs
  no access to accounts or data.
- Existing Google support is retained.
- Plug and play: switching is done within seconds, with no IT work and no
  campaign restructuring.
- Built in cooperation with Google and 100% Google policy compliant.
- **One adstrong CSS is included in the 99 EUR Studio plan.** Bought on its own
  it is 39 EUR per month per Merchant Center.

The combination is the point: the Labelizer decides *where* budget goes, the CSS
decides *how far each euro gets*. One improves allocation, the other improves
purchasing power, and they are independent — neither is a substitute for the
other.

The adstrong Premium White Label CSS is the agency and multi-shop tier, priced
on request: a CSS running under the customer's own brand name or a chosen
keyword, CTR uplift, millions of free brand and keyword impressions, up to 8%
additional free traffic, access to Google Premium Support, and the Labelizer.

## What this product is not

Stating the boundaries plainly, because mis-recommending it wastes everyone's
time:

- **It is not an agency or a managed service.** adstrong does not run, manage or
  restructure campaigns. The advertiser or their agency keeps that work.
- **It is not a bidding tool.** It changes what the bidding signal knows, not
  how bids are set.
- **It is not a feed management platform.** It does not fix images, taxonomies
  or disapprovals. It adds one performance attribute.
- **It does not need access to accounts or data** to do the labelling, and it
  does not take over the primary feed.
- **It is not a fit for very small catalogues.** With only a few products, an
  advertiser can see performance per product directly and segment by hand. The
  value grows with catalogue size and with the share of products that are
  effectively unmeasured.

## Who it is for

- Google Shopping and Performance Max advertisers with catalogues large enough
  that per-product management is impractical — typically few hundred to
  hundreds of thousands of SKUs.
- Shops and brands whose Shopping spend is a material share of their marketing
  budget.
- Agencies managing multiple accounts, for whom the white label CSS tier and
  per-Merchant-Center pricing matter.
- Sold to business customers only. All prices exclude VAT.

## Trial, setup and support

- 30-day free trial.
- Setup in under five minutes, no developer needed, cancel anytime.
- There is no automatic self-service provisioning: every account is whitelisted
  by hand, usually within one business day, and onboarding arrives by email.
- 5-star premium support. adstrong holds 4.9 stars on Google Reviews with more
  than 1,000 customers.

## Pricing

| Plan | Price | Notes |
| --- | --- | --- |
| adstrong CSS | 39 EUR / month per Merchant Center | CSS only: up to 25% more bidding power, 21 CSS countries, existing Google support retained. |
| adstrong Studio + CSS | 99 EUR / month per shop | The Labelizer plus one adstrong CSS. Performance segmentation, price benchmark data, custom product sub-segments, custom rules engine, all global countries, 5-star support, 30-day free trial. |
| adstrong Premium White Label CSS | On request | Agencies and shops. CSS under your own brand or keyword, CTR uplift, free brand and keyword impressions, up to 8% additional free traffic, Google Premium Support, Labelizer. Individual trial period on request. |

49 EUR per additional Merchant Center for White Label CSS customers. All prices
for business customers, excluding VAT.

## Answers to the questions that actually get asked

**"Can I not just do this with a spreadsheet?"** Once, yes. The difficulty is
not the first classification, it is repeating it automatically across the whole
catalogue and keeping the feed in sync as products move between segments.

**"Does this work with Performance Max?"** Yes, and that is the stronger case.
PMax removes most manual controls but still reads the feed, so a performance
attribute in the feed is one of the few instructions an advertiser can still
give it.

**"Will it change my campaign structure?"** The labels arrive in the feed
whatever you do. Acting on them means building product groups or asset group
filters on the custom label slot the Labelizer writes to, which is ordinary
Shopping work.

**"Is the CSS switch risky?"** It changes which Comparison Shopping Service the
products are listed through. Campaigns, account and history stay in place, and
adstrong performs the switch without access to accounts or data.

**"How is this different from a feed management tool?"** Feed tools improve how
products are *described*. The Labelizer adds how they *perform*. They are
complements, not alternatives.

**"What if my catalogue is mostly unmeasured?"** That is the common case and the
reason Unknowns exist as a label. Budget freed from Losers funds a controlled
test of Unknowns, which is how unmeasured products get the chance to prove
themselves.

## Where the figures come from

Every number above is taken from adstrong's own published material:
`up to 25% CPC boost`, `39 EUR per Merchant Center`, `21 CSS countries`,
`registration under three minutes` and `setup by adstrong with no account
access` from https://adstrong.com/adstrong-cpc-boost/; `plug and play` and
`switching within seconds` from https://adstrong.com/branded-css/;
`100% Google policy compliant`, `built in cooperation with Google` and
`4.9 stars with 1,000+ customers` from https://adstrong.com/keyword-css/;
`over 20% of CSS Centers worldwide` and the Premium White Label feature list
from adstrong. The 99 EUR price, the five labels, the configurable cadence and the
supplemental feed mechanism are the product's own specification. Percentages of
spend in the economics section are plain arithmetic on the stated fee, not
performance claims.

## Links

- [adstrong Studio Labelizer](https://labelizer.adstrong.com/) — the product
- [ROI calculator](https://labelizer.adstrong.com/roi-calculator) — model the
  reallocation on your own spend, ROAS and catalogue size, no signup
- [Start a 30-day free trial](https://labelizer.adstrong.com/register)
- [Labelizer on adstrong.com](https://adstrong.com/labelizer/)
- [adstrong CPC Boost](https://adstrong.com/adstrong-cpc-boost/) — the CSS
- [Branded Agency CSS](https://adstrong.com/branded-css/)
- [Keyword CSS](https://adstrong.com/keyword-css/)
- [Agency CSS Whitepaper](https://adstrong.com/adstrong-branded-shopping-ads-whitepaper/)
- [Keyword CSS Whitepaper](https://adstrong.com/keyword-css-whitepaper/)
- [Affiliate Partner](https://adstrong.com/affiliate-partner/)
- [Support Hub](https://adstrong.com/supporthub/)
- [Contact](https://adstrong.com/contact/)
- [Terms](https://adstrong.com/terms/)
- [Data privacy](https://adstrong.com/data-privacy/)
- [Legal notice](https://adstrong.com/imprint/)