Your AI SEO Needs "Value First" Strategies and Reporting

Note from Joe: Please excuse this week’s banner image. I have decided that I hate it. But, I have also decided that I am too tired to make a new one. I promise to get a better image next week. 😔

Since the beginning of SEO, marketing professionals have centered their SEO efforts around metrics such as keyword rankings, organic traffic, and click through rates. For a long time these metrics were a staple to the SEO process because there was a relatively high correlation between them and real business goals. However, because of changes to the search landscape with the inclusion of AI and other SERP changes, and new attribution models in GA4 versus UA, that correlation has gotten worse over the last 5 or 6 years. Now in many cases the correlation no longer exists at all.

One of the most obvious cases is the almost ubiquitous drop across all sectors of Click Through Rates (CTR) after the inclusion of AI Overviews. That has almost completely eliminated SERP impressions as a valuable metric. Rankings have also taken on an almost dubious reliability as Google has aggressively tried to block 3rd party trackers, while the same drops in clicks have decreased the value of rankings in general.

Furthermore, with the advent of Large Language Models, new prompt tracking and brand visibility tools are introducing non standardized metrics whose value rely heavily on how they are implemented by the tools, and articulated to stakeholders. This can leave huge gaps in strategy development and reporting.

While these new challenges to reporting are obvious to most, the bigger problem has been staring SEOs in the face for the last 20 years: Most of the popular SEO metrics like rankings, clicks, and even prompt tracking, have very little direct connection to business goals. Instead SEOs should realign their strategies and reporting to "value first" metrics that are built around top-line business goals.

A metric tree is a hierarchical framework used in data analysis to connect a high-level business or marketing objective to the measurable factors that influence it. It starts with a primary outcome, such as revenue, customer acquisition, retention, or profitability, and breaks that outcome into increasingly specific drivers and supporting metrics. For example, revenue might be decomposed into number of customers × average revenue per customer, while customer acquisition could be broken down further into traffic, conversion rate, lead volume, and close rate.

By showing how individual metrics contribute to broader outcomes, a metric tree helps teams identify which variables have the greatest impact, diagnose performance changes, prioritize growth opportunities, and ensure that day-to-day marketing measurements are tied to meaningful business results.

A useful SEO metric tree should start with business outcomes and work backward toward SEO activity, rather than starting with rankings, clicks, or impressions and trying to justify their business value afterward. The following process can help SEO professionals build a metric tree that makes strategy and reporting more closely aligned with how the business actually measures success.

Start at the top of the tree with the business outcome SEO is expected to support. Depending on the organization, this might be revenue growth, qualified pipeline, new customer acquisition, subscriptions, bookings, or another measurable commercial outcome. Avoid putting organic traffic or rankings at the top of the tree. Those are SEO performance indicators, not business outcomes. The goal is to establish a metric that executives and other business functions already recognize as important.

Break the primary objective into the variables that mathematically or logically influence it. For an e-commerce company, revenue might be expressed as customers × average order value × purchase frequency. For a SaaS business, new recurring revenue might depend on qualified leads × sales conversion rate × average contract value. This layer is important because it establishes where SEO can realistically influence the larger business model without claiming responsibility for outcomes controlled by other departments.

Next, determine which parts of the business equation organic search can meaningfully affect. SEO might influence the number of prospective customers entering the funnel, the quality of those visitors, the products or solutions they discover, and the percentage who progress to a conversion. This creates the bridge between business metrics and SEO metrics. Instead of simply reporting that organic sessions increased 20%, the metric tree can show how additional qualified organic visitors contributed to leads, transactions, subscriptions, or another downstream outcome.

Move another level down the tree and identify the behaviors that connect organic visitors with the desired business outcome. These could include purchases, demo requests, trial registrations, account creation, newsletter subscriptions, product views, or other meaningful actions. Depending on the length of the customer journey, it may also be useful to include intermediate conversions. The important distinction is that these metrics should represent progress toward the business objective rather than engagement for engagement's sake.

Now introduce traditional SEO performance metrics such as organic clicks, qualified organic sessions, non-branded search traffic, landing-page performance, search visibility, and rankings for strategically important queries. These metrics help explain why business outcomes changed. For example, declining organic revenue might be traced to fewer qualified organic sessions, which could then be traced to declining visibility across an important commercial query set. Rankings and traffic therefore become diagnostic metrics rather than the final measure of SEO success.

Further down the tree, connect performance changes to operational SEO metrics that teams can directly influence. These might include indexation, crawlability, internal linking, Core Web Vitals, rendering, content coverage, backlinks, structured data, or other technical and content signals. These measurements are particularly useful for diagnosing problems and identifying opportunities, but they should generally sit toward the bottom of the metric tree. Fixing canonical tags or improving LCP may be important work, but the tree should demonstrate the chain of reasoning connecting that work to visibility, qualified traffic, conversions, and ultimately the business objective.

Once the tree has been assembled, classify the metrics according to their role. Outcome metrics describe what the business ultimately wants to achieve. Driver metrics measure the factors that directly contribute to those outcomes. Diagnostic metrics help explain why a driver is improving or declining. This distinction prevents SEO reporting from treating every available measurement as a KPI. A crawl error count, for example, may be extremely useful diagnostically without deserving the same reporting prominence as organic-generated revenue or qualified leads.

Review each branch and distinguish between metrics SEO can directly influence, metrics it can partially influence, and metrics largely controlled elsewhere. SEO may strongly influence qualified organic traffic but only partially influence conversion rate and have little control over pricing or sales close rates. Making these boundaries explicit produces more credible reporting and encourages collaboration with product, CRO, content, sales, and other teams when business outcomes depend on multiple functions.

The metric tree should ultimately influence what the SEO team works on, not just how it reports results. When evaluating an opportunity, work upward through the tree and ask which performance driver it affects and how that driver contributes to the business objective. An indexing problem affecting high-value product pages may deserve substantially more attention than hundreds of metadata issues on pages with little commercial importance. The metric tree therefore provides a framework for prioritizing work according to expected business impact rather than simply the number of SEO issues discovered.

Finally, structure SEO reporting in the same order as the metric tree. Begin with the business outcome, move into the organic contribution to that outcome, explain the primary performance drivers, and only then introduce diagnostic SEO metrics when they help explain what happened. An executive report might therefore move from revenue → organic revenue → organic conversions → qualified organic traffic → search visibility → specific technical or content drivers. This reverses the structure of many traditional SEO reports and makes the conversation less about what happened in Google and more about how organic search contributed to the goals of the business.

The finished tree might conceptually follow a path such as Revenue → New Customers → Organic Customers → Organic Conversions → Qualified Organic Traffic → Search Visibility → Rankings/CTR → Content, Technical SEO, and Authority. Not every branch will be controlled entirely by SEO, and that is precisely what makes the model useful: it shows where SEO contributes to business performance, where other teams influence the outcome, and which underlying metrics should be investigated when performance changes.

~

SEO does not need more metrics. It needs a clearer understanding of which metrics actually matter. Rankings, clicks, impressions, and AI visibility can still be useful, but they are signals, not outcomes. Their value comes from helping explain whether SEO is contributing to something the business actually cares about. A metric tree makes that relationship explicit. It forces teams to start with value, work backward to the drivers SEO can influence, and use traditional SEO metrics to diagnose what is helping or hurting those drivers. As search becomes more fragmented across traditional results, AI experiences, and whatever comes next, individual visibility metrics will continue to change. The business outcomes behind them will not. The SEO teams that build their strategies around those outcomes will be better positioned to explain their value, prioritize the right work, and adapt when the search landscape changes again.

Neural Network: A model inspired by the human brain made of interconnected processing nodes.

One more thing: AI is only as good as its operator, and if you are reading this newsletter, you’re better than most!

Till next time,

PS: Let me know what you think of this issue, or anything else here: [email protected]