Skip to content

LocalU Advanced August 2021: Practical Local SEO Takeaways

The August 2021 LocalU Advanced sessions gathered local search practitioners to share research, tests and workflows. This guide reorganises the main lessons from four talks into a practical review for local SEOs: how images affect visibility and leads, how to measure content usefulness, how to run local link building, and what to configure in Google Analytics 4.

Images: visibility, ranking and lead value

Joy Hawkins framed image work around three outcomes: looking better in search, ranking better, and generating more leads. The most useful starting point is not metadata but the image itself and where it appears.

For mobile search presentation, the captured guidance recommends avoiding cropped heads, heavy text overlays and stock photography. Square images were preferred, file size did not appear to matter, and placing the image high in the page HTML gave a better chance of it appearing in the search result. Images can also highlight awards or coupons.

Two patterns appeared in the research: product-based images chosen by Google for a query, and cover photos used for the business. Cover photos formed the majority of the results in that research. If a cover photo is not being shown as desired, the captured material points to Ben Fisher's method, with the caveat that some countries, usually in the EU and the UK, show an image result rather than the profile photo.

A useful distinction for readers: geotagging image metadata is not the same as making the image relevant to the page. In the captured tests, geotagging photos uploaded to Google My Business produced no ranking change across 15 tracked keywords. Geotagging photos on business websites also produced no ranking changes where the CMS allowed tagging; a custom CMS stripped the tags out. The practical conclusion was not to bother with geotagging as a ranking tactic.

Adding relevant photos to the linked website page was different. When tire photos were added to a page that did not otherwise mention tires, including in HTML or metadata, a slight local organic ranking increase appeared within three days. The search result used the image alt text as a justification for the business. Replicating this across a couple of clients produced very slight ranking increases. A relevant image with alt text at the top of a lawn-care page also gained a featured snippet and a small ranking bump, but the same result did not appear for a law client.

For lead value, a dentist example showed that adding images to a page that previously had none increased traffic from Google Images. The captured analysis extrapolated that the image-search traffic represented a meaningful share of conversions, but the conversion rate for image-search visitors versus other organic visitors was not established. Treat image search as a channel to measure, not a guaranteed conversion source.

Content usefulness signals: making quality measurable

Rachel Anderson's talk started from a practical problem: expertise, authority and trust are difficult to measure directly, yet search engines rank pages by machine processes. Content Usefulness Signals, or CUSs, shift the question from whether a page feels authoritative to whether specific content differences can be observed and compared across sites.

Examples of usefulness signals include reviews, community information blocks, bulleted amenity lists, parking directions, FAQs and content blocks on eCommerce pages. The method is to isolate the content that matters, extract entities from it, and compare the resulting entity sets and salience scores. Salience indicates how central an entity is to the text.

A patio furniture comparison illustrates the approach. The analysis used only the main SEO text block and category content, excluding products, ads, filters and navigation to avoid clutter. The pages were run through entity extraction, then unique entities were counted and ordered by salience. Although the Lowe's page ranked for more keywords according to SEMRush, the Home Depot page ranked for higher-volume queries and ranked higher, which implies more impressions and a higher click-through rate.

Location pages from big box stores showed a similar pattern. Entity extraction identified which location pages had the richest content. K-Means clustering then grouped topics: Lowe's pages centred on hours, pickup options and addresses, while Home Depot pages included specific products, services and store attributes. SEMRush showed Home Depot ranking for many more keywords on those location pages. The primary differences were store reviews and lengthy content blocks dense with important keywords.

The takeaway is not to copy a competitor's page but to use measurable content differences to decide what to add. For smaller clients, a manual version of the same comparison is often enough: choose the relevant content block, list the entities a competitor covers, and check which of those are missing or underdeveloped on your page.

GA4: what changes and what to configure

Dana DiTomaso's GA4 session treated the move from Universal Analytics as a necessary rethink rather than a simple upgrade. The version history runs from Urchin in 2005, Classic in 2008, Universal in 2013 and GA4 in 2020. Universal Analytics was built around sessions: a session lasts 30 minutes by default, and a visit crossing midnight counts as two sessions. GA4 still has sessions but calculates them differently and records more events, making it more likely to capture session length accurately. Sessions still end after 30 minutes of inactivity, but GA4 is less likely to miss activity on a page where someone spends 50 minutes without a tracked action.

GA4 also handles campaign returns and multiple tabs better. In Universal Analytics, returning to a site through a new campaign, such as a UTM-tagged Google Business Profile link, started a new session attributed to organic search. GA4 does not start a new session in that case.

Everything is an event in GA4, which makes it easier to measure a person's path from start to finish. Bounce rate is gone, replaced by engagement rate. An engaged session is one where the site is open for 10 seconds, a conversion event is completed, or two or more pages are viewed.

Automatic event tracking is enabled by default. Pageviews, 90% scroll depth, outbound clicks, site searches, video engagement and file downloads are collected automatically. Other scroll depths, form fills, tel: or mailto: clicks, non-YouTube videos and other engagement events require custom events. Recommended events provide naming conventions and ideas for what to track.

The GA4 interface changes regularly. In the captured version, the Acquisitions overview had no channels. The Library section allows different reports for different users. Built-in explorations, including a template gallery, are available in the Explorer tab. Path Analysis supports both pages and events and allows choosing the end of the path, so you can see how people reached the contact page. The funnel report is also useful.

For configuration, events can be marked as NPA so they cannot be used for retargeting and personalised ads. Events can be created from other events, such as a 90% blog scroll followed by a newsletter sign-up. Audiences can be built granularly and used to fire events, which is useful for repeat business such as an oil change booking. Goals are now based entirely on events: click the slider to make an event a conversion, but only set goals that relate to actual business goals. A page view is not a conversion. GA4 conversions are more flexible than the 20-conversion limit in Universal Analytics. Conversions cannot be deleted in GA4 but can be switched off.

To set up GA4, you can use the Upgrade button in the Universal property, or create a separate GA4 property and double-tag through Google Tag Manager. GTM is required for GA4 in this workflow. Enable Enhanced measurement, add a GTM trigger using the GA4 measurement ID, then add events. Universal Events need renaming to GA4 Events. Event names should be lowercase with underscores as separators, and everything should be documented.

Debugging uses GTM preview mode or the GA debug Chrome extension. GA4 shows every event as it arrives, and clicking an event reveals the parameters and user properties being sent. Testing is the best way to learn.

A practical review checklist

Use this checklist to turn the session themes into a review plan. It is a decision aid, not a claim that every tactic will work in every market.

Images: check how the business appears in mobile search results. Replace cropped, text-heavy or stock images with clear square photos. Put important images high in the page HTML. Compare the cover photo with the product images Google chooses for relevant queries. Add relevant images with descriptive alt text to the linked website page, not only to the business profile. Measure image-search traffic separately before assuming it converts at the same rate as other organic traffic.

Content usefulness: choose the content block that matters for the query, list the entities a stronger competitor covers, and identify missing or thin topics. Use salience to prioritise central entities rather than counting every mention. Compare location pages for store reviews, services, products and local attributes.

Local links: record every opportunity with source, cost, target page, timeline and contact. Deduplicate link exports from multiple tools. Look outside the immediate competitor set for tactics used by small businesses in other areas and verticals. Follow up quarterly.

GA4: confirm Enhanced measurement is on. List the events that are collected automatically and the events that need custom setup. Rename Universal Events to GA4 Events. Use lowercase and underscores for event names. Mark sensitive events as NPA. Set conversions only for business-relevant actions. Document the measurement plan and test events in debug mode.

Frequently asked questions

Do geotagged images improve local rankings?

In the captured tests, geotagging photos uploaded to Google My Business produced no ranking change across 15 tracked keywords. Geotagging photos on business websites also produced no ranking changes where the CMS allowed tagging. The practical conclusion was not to rely on geotagging as a ranking tactic. Adding relevant images with descriptive alt text to the linked website page is a different test and showed slight ranking movement in some cases.

What are content usefulness signals?

Content Usefulness Signals are measurable content differences across pages, such as reviews, community information, amenity lists, parking directions, FAQs and eCommerce content blocks. The method extracts entities from a defined content block, compares entity sets and salience scores, and uses clustering to identify topics a stronger competitor covers.

Should local link building focus on domain authority?

The captured guidance says not to worry about domain authority or follow versus nofollow for local links. Local links are often relationship-based, hard to reverse engineer and may look unimpressive while still working. A recorded process, client onboarding questions, competitor research and quarterly follow-up matter more than metric filters.

What does GA4 collect automatically?

GA4 automatically collects pageviews, 90% scroll depth, outbound clicks, site searches, video engagement and file downloads. Other scroll depths, form fills, tel: or mailto: clicks, non-YouTube videos and other engagement events require custom events. Enhanced measurement should be enabled during setup.

What counts as an engaged session in GA4?

An engaged session is counted when the website is open for 10 seconds, a conversion event is completed, or two or more pages are viewed. GA4 replaces bounce rate with engagement rate.