Six Query Families in a Sourcing Account: What to Check Before You Add a Negative
Brand promotion: this post introduces Guangsuan's Google Ads management service for cross-border sourcing advertisers. It is a sponsored vendor introduction, not an independent review, a customer story, or a commissioned evaluation.
If you run Google Ads for a sourcing, procurement or supplier-facing business, your search terms report will contain the buyers you wrote the campaigns for, and everyone else. The second group is specific. Suppliers pitch their own factory (“we can supply X, contact us”). Recruiters advertise roles. Downloading sites offer free supplier lists, audit templates and import guides. Consumer questions arrive about buying one unit for a home workshop, and second-hand listings appear for equipment you only supply new. None of these are stupid searches, and a shared negative keyword list is a blunt instrument: it blocks the two-sided marketplace and the small-order customer along with the job seeker.
This article answers one question for that reader: how should a mixed search terms report be worked through, and which parts of the work does a Google Ads management service actually take on?
The short answer
Answer two questions about each family of query before you exclude anything: could this searcher become a customer, and does the page behind the ad answer what they just asked? Then work in this order.
- Group the visible queries into families by what the searcher was trying to do — buying, evaluating, or something the business does not serve.
- Write one decision per family: keep, watch under a stated condition, or exclude — with a business reason that would still hold next month.
- Carry the decision out with the narrowest negative type and scope that solves it.
- Adjust bids and budgets only after the record exists. A bid change sets a price on exposure; it does not change who typed the query.
Two conditions come with that answer. The order is written for search campaigns aimed at buyers who are actively looking for a product or a supplier; a reach or awareness format, where one click is not expected to become an enquiry, only partly follows it. And the deliverable is a short table another person can pick up, not a feeling that the account is tidier than it was on Monday.
What the report can and cannot tell you
The report shows the queries that matched your keywords and triggered an ad, which makes it a record of what the budget actually bought. It is not a complete inventory of everything the account matched. Guangsuan's published guidance on the search terms report states that queries with insufficient activity can be withheld under data privacy standards, and that the related insights view groups low-volume queries under a subtheme without revealing the original wording. Read the visible rows as a sample you can act on.
The insights view carries its own trap. Its category and subcategory labels summarise a group of searches, and a label is not necessarily one of the queries inside it. An unwelcome label is therefore a prompt to read the visible queries and the dates they ran in, not a conclusion that a negative has stopped working.
Two practical notes before you compare anything:
- Exports do not widen the data. Guangsuan's guidance is explicit that export tools and third-party services save time on organising the data but still return what the platform permits.
- Align the scope before comparing two periods. Record the date range, campaigns, filters, currency and conversion actions, and check whether paused campaigns or removed keywords are included. Two people quoting “this month's spend” may be looking at different sets of rows.
Six query families to work through
Guangsuan's guidance puts the test this way: describe what the searcher is trying to do in full, then check it against the products, order sizes and services the business can actually deliver. For a sourcing or procurement account, six families give a workable starting structure.
| Family | What it looks like in a sourcing account | Ask before excluding |
|---|---|---|
| Buying | OEM supplier, factory direct, custom manufacturer, MOQ, wholesale price or lead-time terms from a stated market | Does the ad lead to the page that answers this term, rather than to the home page? |
| Evaluating | Selection and sizing questions, comparison requests, requests for a drawing, a certificate or a factory audit | Is this research that leads to an order, or a specification exercise with no purchasing step? |
| Suppliers pitching you | “We are a factory”, “we can supply”, “our price is $x”, a supplier introducing their own plant | Is this account buying or selling? If you buy, or you sell data rather than goods, is a supplier pitch noise or a lead? |
| Job and salary searches | Recruiter ads and CV submissions for roles you do not have | Does any part of the business recruit, at any location, in any language? |
| Free resources | Supplier-list downloads, import guides, factory audit templates, sourcing spreadsheets | Does a free resource count as a lead for you, or does it absorb budget you would rather spend? |
| Single units and second-hand goods | “One piece”, home workshop quantities, DIY or how-to-make queries, used equipment and spare parts | Do you take retail-sized orders or second-hand stock? The answer is a business rule, not a keyword. |
The third family is worth pausing on, because it exists because of the business model rather than the query. A sourcing platform or a procurement team attracts suppliers as well as buyers. If you are on the buying side of the transaction, supplier pitches are not leads and can be filtered; if you run ads for a marketplace, a supplier registration flow or sourcing advisory work, those same queries may be worth keeping. The family is identical in both accounts. What decides the action is which side of the trade your business sits on — and that assumption is the kind that quietly changes when a company adds a new service line, so it belongs in the account notes.
The sixth family carries a trap in the other direction. A parts query is noise for a project supplier and core business for a manufacturer that keeps a parts desk. Excluding “parts” at account level because of one product line blocks demand you may have confirmed with a customer.
Turn each family into a written decision
The output of a review should be a table someone else can pick up: the query or group, the keyword it matched, the family, the decision, the negative type, the scope, the business reason and the date. Two habits make that record worth keeping.
First, write the reason as a standing rule rather than a note about this month. “We only take container-scale orders”, “we cannot ship to that market” and “we do not run a repair service” are stable; “no conversions last week” is an observation that expires and is not a safe basis for a permanent block. Second, ask the sales team to record a specific reason behind every unusable enquiry — job applications, sales approaches, products you do not sell, markets you cannot deliver to, order sizes you do not take, or lead times that cannot be met — and keep “not yet contacted” separate from “confirmed unsuitable”. Otherwise a follow-up problem will be read as a traffic problem and a whole family will be excluded for the wrong reason.
Keep a second short list next to the decision table: the demand you must not block. A sourcing business still takes orders of a size that consumer phrasing never suggests, and a blanket negative on a word that represents an entire line of business will quietly remove those orders.
Choosing the negative type and scope
Guangsuan's guidance on negative keywords notes that negatives do not expand to close variants the way positive keywords do, so singular and plural forms, alternative spellings and translated terms may each need their own entry once the report shows them arriving. It also notes that negatives handle case changes and the spelling corrections Google makes for you, which is not the same as excluding every related variant. And maintaining a list is not importing a large dictionary: the intent being excluded, the match type, the scope and whether other related word forms will still get through all need confirming.
Scope is the other half. A family you do not serve in one product line does not justify an account-level block on the whole category, and Guangsuan's guidance is to pick the smallest scope that solves the problem and keep the previous setting, so an exclusion that turns out to be too wide can be reversed. There is also no need for a weekly quota of new negatives: the demand that should be excluded is already excluded, and a larger list may only block business that had already been confirmed.
Bids come after the record
When the complaint is wasted spend, three causes produce the same symptom, and only one of them is fixed inside the account: the traffic is not the audience you want, the page is not holding the visit, or the data is not clear enough to tell which. Guangsuan's guidance describes the same three inputs in its own terms — visible queries tell you which intent has entered the account, the landing page shows whether that demand can be met, and sales records tell you what actually came in — and its published service path follows the same sequence: identify purchase intent and exclude irrelevant demand, send the search term to the relevant product page rather than routing every click to the home page, then use the form, the contact details and the sales response to steer the next round. The stated basis for ongoing optimisation is three inputs read separately: Google Ads account data, website analytics and sales feedback.
One more caution about the metric itself. A lower share of unmatched spend is not automatically good news: if the visible queries are products the business does not supply, more visibility will not make the enquiries better suited. And the report cannot be bought open — Guangsuan states that queries withheld under data privacy rules cannot be released by raising bids, adding clicks or splitting ad groups.
What Guangsuan's Google Ads management covers
Guangsuan describes its Google Ads management as running from search terms and ads through to landing pages and conversion tracking, with optimisation aimed at real purchasing demand rather than clicks alone. Its published scope lists search terms report analysis and negative keyword maintenance as separate items, alongside account structure and campaign setup, budget, area and bid optimisation, core and long-tail keyword research, landing page analysis and URL path matching, website analytics linking with tracking and conversion setup, creative optimisation, conversion performance and ROI analysis, and periodic reporting with review meetings and follow-up actions. Items are combined per project, and material production, page development and data integration are confirmed in the plan and contract rather than assumed.
The fees are published separately from the ad budget, which is worth understanding before the first quote. The service page lists a one-time account setup fee of ¥2,500 for the first opening and initial setup, a service fee of 15% calculated on each ad top-up, and a monthly minimum operating fee of ¥1,500 when monthly ad consumption falls below ¥10,000. Guangsuan states that these published figures do not represent a promise of enquiry volume or return on investment, and that the settlement and offset between the top-up service fee and the monthly minimum, and the treatment of a paused account, are confirmed item by item before signing rather than assumed.
Two limits are published in the same place. Guangsuan states that it does not promise a fixed enquiry volume or a fixed deal outcome, and that ad formats, placements and billing follow Google's current rules and the features available in the account. The company also states that it was founded in 2019 and works on a corporate contract with a defined delivery scope. The scope, the fee structure and the limits sit together on the Guangsuan Google Ads management page, so they can be read as a checklist rather than a promise.
What to bring to the first review
Guangsuan's published starting list for an existing account is short: the website address, target market, the current difficulty, and the historical data you can authorise. Account access, negative keyword changes and tracking adjustments are agreed separately, so a first discussion can begin with redacted material and a diagnosis of the account structure, the search terms, the conversion setup and budget performance.
Four additions turn it into something a sourcing team can act on:
- The queries from the last quarter that sales would recognise as real buyers, and the ones they would not.
- The order sizes and markets you do not take, so each family's business reason is a rule rather than a guess.
- What the account currently counts as a conversion, and what your team counts as a qualified enquiry.
- The same-period search terms export, campaign summary and current negative keyword list with their match types, with names, emails and phone numbers removed.
If the review is going to end in a reconciliation against CRM or finance records, the export conventions have to be settled before the numbers are compared. Guangsuan's guide to exporting Google Ads data to JSON covers the field names, the date and time zone basis, and the currency and conversion handling that a monthly close depends on.
One check comes before the account, not after. Guangsuan's position is that a site can be evaluated before ads go live, but that skipping the basic checks is not advisable: the product and target market need to be clear, the site should open normally, and product information, contact routes and conversion tracking need to be able to receive traffic.
Frequently asked questions
Our searches attract suppliers pitching their own factory. Should we exclude all of them?
Only if the account is on the buying side of the trade. A sourcing platform selling data or advisory, or a marketplace, may want those queries, so decide per campaign group rather than account-wide, and record which side of the business each group serves.
Can we import a standard negative keyword list to save time?
Use it as a starting point, not as the method. Guangsuan's guidance is to confirm the intent being excluded, the match type, the scope and whether other related word forms will still get through, rather than importing a large dictionary — because a shared list was written for someone else's supply range and market.
Should we exclude every query that did not convert?
No. A relevant query can fail to convert because the page, the offer or the follow-up was the problem, and excluding on a quiet week removes the evidence you would need later. Record a standing business reason instead.
Do negatives cover plural forms, misspellings and other languages?
Not automatically. Negatives do not expand to close variants the way positive keywords do, so variants may need separate entries once the report shows them arriving. A query written in another language belongs to the same family as its English equivalent; the vocabulary changes, the decision does not, so read the actual queries rather than a translated list.
What if excluding a family reduces enquiries as well as clicks?
Fewer clicks is not a success signal on its own. At review time, check whether the mismatch you targeted has fallen, whether the demand you wanted to keep is still arriving, and how qualified enquiries relate to spend now. If qualified enquiries fell, the exclusion is probably too wide or the page has weakened.
Where Guangsuan fits
Guangsuan provides Google Ads management for export and B2B enquiry generation, covering the search term review, ad structure, landing page matching, conversion setup and sales feedback described in this article. The published scope, fee structure and stated limits are on one page, so a sourcing team can read them as a checklist before deciding whether the service covers the part of the review it does not have time to run.