A measurement that allows defining specific metrics relevant only to business goals, moving past generic industry standards.
People concerned with brand visibility in complex AI search answers, reading it alongside standard KPIs.
01What It Is and How It Works
Standard metrics (like overall click-through rate or total impressions) provide a broad view of performance. However, they often fail to capture the nuances of how your brand is perceived within complex AI search answers. A Custom Metric solves this by allowing you to build proprietary calculations based on raw data points that matter only to your industry. For instance, if your goal is establishing authority, simply tracking impressions isn't enough; you might need a metric that weights 'mention frequency in synthesized summaries' against 'direct citation count.' The system processes the underlying search interaction data—the text snippets, the source attribution, and the context of the surrounding AI-generated content—and applies your specific formula to generate a single, actionable score. This moves measurement from what happened (a standard metric) to what it means for us (your custom metric).
If standard metrics like clicks or impressions don't tell you what matters for your brand right now, a Custom Metric lets you build your own scorecard. You define exactly what success looks like for your brand in AI search results and then track that specific number over time.
02What To Do About It: Concrete Actions This Week
Do not build a custom metric simply because you can. The process requires strategic alignment first. This week, gather stakeholders from marketing, product development, and sales. Present them with three core business questions that the current reporting dashboard cannot answer regarding brand visibility in AI search. For example: 'Are we appearing as the definitive source for X topic?' or 'Is our competitive differentiation (Y feature) being highlighted when we are mentioned?' Once you have consensus on these high-level gaps, work with your analytics team to map those questions back to specific data inputs available from the platform. Your goal is not to track more numbers; it is to create one single metric that answers the most critical strategic question."
03How It Is Measured or Noticed
When you implement a Custom Metric, it will appear alongside your standard KPIs within the platform's reporting interface. Instead of seeing 'Total Impressions,' you might see a new card labeled 'Authority Score (Custom)' or 'Synthesized Mention Ratio.' The key thing to look for is the breakdown provided beneath the primary number. Good custom metrics will not just give you a score; they must show you which underlying data components contributed most heavily to that score—for example, showing that 60% of your current Authority Score comes from 'direct citation in summary' and only 15% comes from 'snippet appearance.' This breakdown is crucial because it tells you where the measurement system is currently finding success or failure.
04Common Mistakes When Defining Metrics
Defining a custom metric is powerful, but it is easy to get bogged down in vanity numbers or overly complex formulas that are impossible to action. Always keep the business outcome at the forefront of your definition.
- Creating metrics based on correlation rather than causation. Just because two things move together doesn't mean one is causing the other; ensure your metric tracks a factor you can actually influence. — warn
- Making the formula too complex. If stakeholders cannot understand or explain how the score is calculated in three sentences, it is too complicated to be useful for weekly decision-making. — warn
05Limits and Confusion: What It Is Not
A Custom Metric is a powerful diagnostic tool, but it must not be mistaken for an absolute truth or a direct ranking factor. First, remember that while the metric tracks your appearance in AI search results, it does not guarantee future performance; human behavior and algorithm updates are always variables outside of its scope. Second, do not confuse the Custom Metric score with organic SEO rankings (like traditional Google SERP positions). The two measure different things: one measures visibility within a specific format, while the other measures positioning in a list. A high custom metric suggests strong brand presence and trust signals, but it requires complementary strategies to improve actual click volume.
06A Worked Example: Measuring Trust Signals
Imagine a B2B SaaS company whose primary goal is being seen as the most reliable industry expert. Standard metrics might show they rank highly (good), but they don't measure why. The team defines a Custom Metric called 'Expert Authority Index.' This index formula weights three inputs: (1) Number of times the brand is cited by third-party sources within AI summaries, (2) Average length of text attributed to the brand in those summaries, and (3) Frequency of mention alongside competitor names. A high score means the AI isn't just listing you; it's actively using your content as foundational proof points for its answer.
The goal shifted from 'How many times are we mentioned?' to 'Are we being used as the definitive source of truth in AI answers?'
Frequently asked questions
How is a Custom Metric different from just tweaking an existing standard KPI?
It differs because standard KPIs measure broad, industry-level performance signals like overall impressions. A Custom Metric allows you to define a measurement that captures a highly specific business objective unique to your brand's goals, such as quantifying the perceived reliability of information in niche search queries.
Should I create custom metrics for every goal I have, or is there a point where they become useless?
You should only build a Custom Metric when you can translate an abstract business objective into a measurable, actionable data point. If the metric requires overly complex formulas or measures something that cannot directly inform immediate content changes, it is likely a vanity number and should be avoided.
If I implement a custom metric, how quickly will I see results in the reporting dashboard?
The data for a Custom Metric begins appearing in the platform's reporting interface as soon as you successfully implement it. However, generating meaningful trends requires sufficient search volume and consistent data input over time; initial readings may be noisy until enough samples accumulate.
Can I use custom metrics to prove that my brand is definitively ranking higher than a competitor?
No, you cannot mistake a Custom Metric for an absolute truth or definitive ranking factor. It is a powerful diagnostic tool that helps quantify how users perceive your brand relative to its goals (like trust), but it does not provide the raw algorithmic ranking data itself.
If my custom metric shows a sharp drop, what should I check first before assuming a campaign failure?
First, verify that the underlying mechanism tracking the metric is functioning correctly and hasn't been impacted by platform updates or changes in search behavior. Next, cross-reference the dip with external factors, such as major industry news cycles or competitor launches, to determine if the drop is systemic rather than performance-related.
Asked out loud
spoken, not typedThe same term in the words somebody uses speaking to an assistant rather than typing into a box — written from the situation, which is why each one carries the situation it came from.
You should define a Custom Metric that specifically quantifies the signals related to expertise and reliability within your industry context. This moves you past generic metrics like total clicks and focuses only on the data points that prove your authority, making it actionable for content creation.
It depends on your primary business goal; you shouldn't focus on all the numbers provided. Instead, identify the single most critical objective—like proving thought leadership—and build a Custom Metric around that specific measurement to filter out the noise and guide your strategy.
Yes, you can move beyond simple volume metrics by building a Custom Metric that tracks qualitative signals relevant to your brand's goals. For instance, instead of tracking total clicks, you could track the frequency with which users engage with 'how-to' guides on your site after seeing an AI search result.