12 questions to ask a brand-search monitoring vendor before buying

A useful vendor demonstration leaves you with answers that can become acceptance criteria. Send the questions below before the meeting and ask the presenter to use your proposed brand, markets and devices. Keep a record of which answers apply to the quoted package and which describe a separate module or planned feature.

Ads Lighthouse publishes this checklist as a vendor in the category. These are questions we think buyers should ask us too. They do not assume that every team needs every capability; an explicit limit is easier to plan around than an ambiguous promise.

1. What exactly counts as one check?

Ask the vendor to calculate usage for a sample configuration while you watch. Include multiple keywords, two locations and both devices. Establish whether engine, language or operating system adds another billable combination, and whether retries or manual refreshes consume the same allowance as scheduled collection.

2. Which parts of my coverage can you actually collect?

Provide the exact locations and search engines you need. Request a written list of supported combinations and exclusions. A statement about worldwide coverage does not specify whether your purchase includes the city, language, device and ad format that matter to your team.

3. What timestamp am I seeing?

Have the presenter distinguish observation time from ingestion time and the time the page was opened. Ask which timezone a report uses. If a result is cached, delayed or retried, the original collection context should remain understandable to someone reading the finding later.

4. How do you display an unsuccessful check?

Request a failed-check example. Find out how you distinguish an empty result from a collection failure and how long incomplete work remains pending. Ask whether the service reports coverage gaps, who investigates them and whether the contract provides any remedy for persistent gaps.

5. Which URL are you preserving?

Google describes display URLs, final URLs and tracking templates. Ask the vendor to identify every URL field it exposes and where each value came from. Establish whether it saves redirect information, what can be missing and whether reopening a destination fetches a different page from the one originally observed.

6. What does your visibility metric divide by?

Ask for the denominator in plain language. Is the metric based on captured ad results, checks containing an advertiser, eligible account impressions or a model? Request the calculation for one small example. Keep the definition beside the report so colleagues do not treat unlike percentages as interchangeable.

7. Can we classify an advertiser without accusing it?

Demonstrate a competitor, an authorized retailer and an unresolved domain. Ask whether relationship labels are distinct from findings about particular behavior. Check who can correct a label and how a changed relationship affects historical reports. Permission can differ by market or date.

8. Which actions happen automatically?

Trace a new observation through collection, detection, notification and any external response. Ask where approval is required and who can turn automation off. A system that supports email delivery may still require someone to create the event that triggers an email; test the entire path.

9. Who can access each brand's information?

Test with an account that should not see another project's data. Include direct links and exports, not only navigation. Establish account ownership, user removal and any limits on client access. Agency buyers should make project isolation a demonstrated requirement before inviting unrelated clients.

10. What can we take with us when we leave?

Request a sample export and open it before signing. Check whether it contains original timestamps, query context, domains and useful identifiers. Ask about retention after cancellation, export charges and whether attachments require a separate retrieval process. Write the agreed exit process into the commercial record.

11. What changes the monthly cost?

Price your normal configuration and a busy month. Include additional brands, manual checks, more frequent collection, users and service work where relevant. Confirm what happens at a limit: collection stopping, a queued request or an additional charge are materially different outcomes for the team.

12. What will prove the trial succeeded?

Agree practical tasks before the trial begins, including one uncertain finding and one correction. Name the reviewers and the decision date. End with a list of demonstrated capabilities and unresolved requirements, and ask the vendor to confirm it. Use that record alongside the price when deciding whether to proceed.

For an Ads Lighthouse evaluation, bring these questions to our contact page. Our current product focuses on configured observations and manual review; the evaluation should show that actual workflow and state any requirements it does not meet.