Playbook for Quick Wins with Product Data

0. What this is for

Shopping / P-Max match your products to queries using the data from your feed. A properly written title and a meaningful description directly raise relevance and the number of impressions. The data requirements are described in the Product data specification.

Click share shows how large a part of the potential clicks you are currently taking — a great indicator of whether there is room to grow.

Price affects both impressions and conversion: in Merchant Center there is a Pricing / Price competitiveness tab where you see where you are more expensive than the market and by how much.

1. Feed check

I know it sounds very Generic, but it needs to be checked. Both access to the auctions and traffic quality rest on it.

1. Check the required attributes per the specification: id, title, description, link, image_link, price, availability, condition, brand/gtin/mpn, google_product_category and so on. Formats and language according to the target country.
2. GTIN: add it everywhere it exists — it helps Google match products correctly and unlocks features.
3. Images: clean, large, without watermarks/frames — a violation leads to warnings/disapprovals. (See Product data spec.)
4. Price and availability sync: turn on Automatic item updates — Merchant Center temporarily fixes price/availability using the data on the page when the feed hasn’t managed to update. It is not a replacement for regular updates, but a great safety net.
5. Update the feed regularly (scheduled fetch / Content API) and keep Diagnostics in MC clean.

2. Where to click: let’s take the right reports (Google Ads + GMC Next)

2.1. Customer language → Search terms

• Path: Google Ads → Campaigns → Insights & reports → Search terms → period of the last year.
• Columns: Impr., Clicks, CTR, Cost, Conversions, Conv. value/ROAS, Avg. CPC.
• Goal: find out what words people really use to search for your product and which phrases bring sales.

2.2. SKU level → Shopping products (Item ID)

• Path: Google Ads → Reports → Predefined (Dimensions) → Shopping products.
• Columns: Impr., Clicks, CTR, Click share, Cost, Conversions, Conv. value/ROAS, Avg. CPC.
• Goal: find the “mute” items (low CTR), the “pulls but doesn’t sell” ones (high cost, weak conversions) and the bestsellers.

2.3. Traffic potential → Click share

• Where: add Click share to the Shopping/P-Max campaign tables.
• Analysis: low Click share at a normal CTR = you are not taking the available clicks (often bidding/competition); low CTR = a relevance problem (title/image/price).

2.4. Price vs market → Merchant Center → Analytics → Pricing

• Path: Merchant Center Next → Analytics → Pricing (the price benchmarks and optimal sale prices cards).
• Goal: find out where you are more expensive than the market (by GTIN) and which SKUs are dragging efficiency down.

2.5. What the SERP looks like → Ad Preview & Diagnosis

• Path: Google Ads → Tools → Ad Preview and Diagnosis, or enter the Search Terms into Google manually and look at who shows up and with what descriptions.
• Goal: look at competitors’ tiles on key queries without inflating impressions.

3. Prioritisation: what to tackle first (3 rules)

1. Low CTR at normal impressions → a weak/irrelevant title (and/or image). Move the “customer language” from Search terms to the beginning of the title; specify type/model/material/who it is for.
2. High cost, few sales → the demand is there, but the price/page isn’t convincing. Check Price competitiveness and the product page (photo/description/benefits).
3. You are 5–10 %+ more expensive than the market on competitive SKUs → a high ROAS is hard to expect; flag these items and cut the budget on them, or gradually adjust the pricing policy. (Google is moving the reports to the new Price competitiveness module.)

4. Product Title (formulas, examples, what to put first)

The basics: write in the buyer’s language, put the most important thing at the beginning; length up to ~150 characters; no promo words and no CAPS; the title must match the landing page.

4.1. When the brand is strong (there is brand demand)

Pattern: Brand + Model/Type + key feature + size/colour/material + use-case

Example: LG 42LA690 42″ 4K 120 Hz TV – ThinQ AI, HDR10

Recommendation: “front-load” the key info for better relevance

4.2. When the brand is unknown / private label

Pattern: Type + key feature + dimension/material + use-case + (Brand at the end/optional)

Example: Pot 28 cm stainless steel – induction, oven-safe, with lid – [Brand]

The key: start with the words people use to search for the product

4.3. Templates for the main verticals

• Home/kitchen: Brand + Type + Material/Diameter + Compatibility/Use-case + (Number of pcs)
• Clothing/footwear: Brand + Product + Women/Men/Kids + Colour + Size + Fit/feature
• Electronics: Brand + Model + Size/Spec (4K/120Hz) + Type + key feature

Based on best practices for feeds and industry examples

4.4. The story of “one word = ×4–×6 the demand”

For two years we called the product “ice cream makers”, but people searched for “popsicle mould” 4–6× more often. We rewrote the title on all SKUs — impressions, clicks and orders grew, and the CPC went down.

The point: write in your customers’ words, not internal slang; and start with exactly those words.

5. Product Description (fast and human)

• The first 150–500 characters: the most important part — who it is for, the main benefits, compatibility/material/packaging.
• Then the details: features, answers to objections, care/warranty, usage scenarios. Write as if you were explaining the product in a shop (not just dry parameters).
• The description must match the landing page and the policies; formats according to the specification.

6. Price: where to look and how to decide

• Merchant Center → Analytics → Pricing (Price competitiveness) shows where you are more expensive than the market and gives tips for a promotional price. That affects clicks, conversions and revenue.
• On SKUs with high demand but weak conversion, test cutting the price by 3–4 % and check every 1–2 weeks.
• On rare/niche SKUs (bought a couple of times a year) +3–4 % can work (monetising the “scarcity”).
• The older “price benchmarks” are being replaced by the new Price competitiveness report.

7. Rolling it out without developers: two ways to change the title

7.1. Attribute Rules in Merchant Center

• Turning it on: Settings & tools → enable Advanced data source management, then Products → Feeds → Primary → Rules.
• How to do it: build a rule for title (or structured_title), “glue together” brand / type / colour / size / material and so on. Check Preview and Apply.

7.2. Supplemental feed

• Adding it: Settings & tools → Data sources → + → Supplemental feed.
• What to upload: CSV/Google Sheets with only the fields you need (id, title or structured_title, custom_label_*, anything you want to change and try).
• Linking: by id to the primary feed — you change only the selected items, without touching the source.

Leave Automatic item updates on as a safety net for price/availability swings.

8. Fair A/B tests of the Title (without duplicate SKUs)

1. In the master feed make a random 50/50 split: custom_label_0 = title_A and title_B.
2. Apply the new title formula only to group B via a Supplemental feed or Rules (do not change the ID!).
3. Keep the same campaigns; compare A vs B by: CTR, Click share, Avg. CPC, Conv. rate, CPA/ROAS. Test for 2–4 weeks or until there is a clear difference.
4. Do not duplicate a product with a new ID for the test — you’ll create self-competition and “noise” in the data.
5. Roll the winner out to the whole category.

9. “Where to click” — the short version

1. Google Ads → Search terms: export the phrases that carry conversions/value.
2. Google Ads → Reports → Predefined → Shopping products: add CTR, Click share, ROAS and pick out the SKU candidates.
3. Merchant Center → Analytics → Pricing: check where you are more expensive than the market.
4. Ad Preview & Diagnosis: look at how the competition builds titles on your queries.
5. MC → Rules / Supplemental feed: build the new title formula (most important first).
6. A/B via custom labels: run the test on a part of the catalogue (test by product category); decide after 2–4 weeks.

10. A 14-day plan

Days 1–2 — Analysis

• In Ads: download Search terms and Shopping products (Item ID) for 30–90 days.
• Flag 20–50 SKUs with a low CTR and/or high spend at a weak ROAS.

Days 3–4 — Formulas and labels

• Prepare title templates for 2–3 categories (brand / no-name / skeletons for the verticals).
• Set custom_label (A/B), prepare a Supplemental feed or Rules.

Day 5 — Start

• Apply the new titles to group B; check in MC Diagnostics that there are no errors.

Days 6–11 — Monitoring

• Watch CTR, Click share, Avg. CPC, Conv. rate, ROAS weekly for A/B; check Pricing and Diagnostics.

Days 12–14 — Decision

• Confirm the winner; roll the formula out to the category; plan the next test (attribute order, a new key feature, synonyms from Search terms).

11. Frequently asked questions

Does the brand always have to be at the start of the title?

When the brand is well known and has brand demand — yes, it usually raises relevance (“front-load”). When the brand is unknown — start with the type and the key features, and the brand can happily go at the end. (In line with the recommendations for feeds and with practice.)

Where will I see the impact of price on performance?

GMC Next → Analytics → Pricing (Price competitiveness) — a comparison of your price with the market and suggestions for promotional prices.

How do I look at the results without inflating impressions?

Through Ad Preview & Diagnosis — a safe preview by query/geo/device.

Should I turn on Automatic item updates?

Yes, as a safety net against price/availability mismatches with the page. But regular feed updates are still necessary.

12. A mini cheat sheet for the Product Title

Strong brand: Brand + Model/Type + key feature + size/colour/material + use-case

Example: Tefal Pan 28 cm Titanium – induction, oven-safe, set of 2

Private label / no-name: Type + key feature + dimension/material + use-case + (Brand)

Example: Midi dress – viscose, beige, size M – [Brand]

Clothing: Brand + Product + Women/Men/Kids + Colour + Size + Fit/feature

Electronics: Brand + Model + Size/Spec (4K/120Hz) + Type + key feature

Home/kitchen: Brand + Type + Material/Diameter + Compatibility/Use-case + (Number of pcs)

13. What to measure and how (after the changes)

• CTR ↑ and/or Avg. CPC ↓ in the test group = better title relevance.
• Click share ↑ = you are taking a larger share of the available clicks in Shopping/P-Max.
• CR/CPA/ROAS = the final check that it is not only getting clicked but also paying off.
• Pricing (MC) = you are not systematically more expensive than the market where you want to scale.

14. Useful help pages (save them to your notes)

• Product data specification (attributes, formats, limits): Google Merchant Center Help.
• Search terms (where to find the report): Google Ads Help.
• Click share (what it is and how to work with it): Google Ads Help.
• Ad Preview & Diagnosis: Google Ads Help + anonymous preview.
• Automatic item updates (how it works and why): Merchant Center Help.
• Pricing / Price competitiveness (MC Next → Analytics): Merchant Center Help.
• Attribute Rules / Supplemental feeds (how to change titles in bulk quickly): Merchant Center Help.

The final thought

The Product Title is your “SEO” for Shopping, the Description helps both the algorithm and the customer, and Price is the bridge to conversion.

Write it in people’s language, check the data regularly and test in small iterations — you’ll see impressions, clicks and orders grow without magic and without complexity.