It's an impression count dressed up as a pipeline signal. A contrarian case for killing SoV as your AEO KPI — and what to measure instead.
Share of voice was always a branding metric. AEO tools rebranded it as a pipeline metric, and it's been costing B2B SaaS CMOs their renewals ever since. If you're going into a 2026 CFO review with a share-of-voice chart as your headline number, you're walking in with evidence you don't know what your AEO tool actually did for the business.
This is the contrarian essay. Pipeline, not share of voice. The architectural reason SoV fails as a KPI. And the replacement metric most B2B SaaS CMOs should have been tracking all along.
Walk back through the history. Share of voice in television meant: of the ad minutes running in your category, how many were yours. Share of voice in print meant: of the column inches in your category, how many mentioned your brand. Share of voice in Google meant: of the search-result real estate for your keywords, how much you owned. In every case, SoV is a denominator — a count of total impressions you claimed from the total impressions that were there to claim.
AEO tools imported SoV without questioning whether the denominator still made sense. Of the AI answers generated for these prompts, how many cite you. It's a clean definition. It's also an impression count. An answer engine citing you is an impression on the buyer, not a transaction with the buyer.
Impressions have never been pipeline. A pre-roll YouTube ad for Salesforce does not generate a HubSpot deal. A billboard on 101 does not generate a Notion ARR number. And a ChatGPT citation showing Acme Corp for "best CRM for B2B SaaS" does not generate a sales opportunity — unless the person asking clicks, lands, self-identifies, and enters your pipeline. That last step is attribution. The first step is awareness. Share of voice only measures the first step.
"Monitoring, not attribution" — the one-line takedown that surfaces, unprompted, in nearly every CMO discovery call we run. That's the whole essay. The rest is math.
SoV rewards breadth. Pipeline requires intent. The two metrics pull in opposite directions.
If your tool measures share of voice across 500 prompts, you grow the number by showing up for more prompts. The math is incentive-aligned with casting a wider net — more prompts, more citations, bigger SoV number. But a random extra prompt is much more likely to be low-intent (generic, informational, comparison-shopping at a stage of the funnel that doesn't close) than high-intent (evaluation-stage, shortlist-making, purchase-adjacent). You can grow SoV by 40% without growing pipeline at all.
Pipeline works the other way. Pipeline requires you to be cited for the specific prompts that high-intent buyers ask when they're actually shopping. In most B2B SaaS categories, that's 7–12 prompts total, not 500. Those prompts are usually competitive ("X vs Y"), alternative ("alternatives to X"), specification-heavy ("best X for B2B SaaS with 100 seats on HubSpot"), or comparison-heavy ("is X worth it"). The rest are discovery-phase or off-intent.
A CMO optimizing for share of voice is optimizing to show up in all 500 prompts. A CMO optimizing for pipeline is optimizing to dominate the 7–12 prompts where buyers actually decide. The two strategies diverge fast. After six months, an SoV-maximizing strategy has broad mediocre presence across the category, and a pipeline-maximizing strategy has near-total coverage on the 12 prompts that matter. The second one closes deals. The first one shows up nicely in a screenshot.
The question to ask your AEO tool this week. Of the prompts we're tracking, which ones produced a contact that entered pipeline? Which produced a contact that closed? If the tool can't answer, it's measuring the wrong thing. It's monitoring, not attribution.
In the audience brief we compiled across 2025–2026 sources, the phrase "share of voice went up but pipeline is flat" showed up in 14 of 20 CMO discovery calls without prompting. The pattern:
This isn't because AEO doesn't work. AEO in 2026 is measurably producing pipeline — Webflow's 6x ChatGPT-conversion number, Vercel's 10% of signups from ChatGPT referrals, Backlinko's 800% YoY LLM-referral growth are not imaginary. AEO works. The SoV metric just fails to capture it. A tool that reports SoV can be honestly doing its job while the channel underneath is generating pipeline it cannot see.
Three metrics. One of them replaces share of voice entirely; the other two provide context.
The headline metric. Sum of (deal amount × stage probability) across all deals in your CRM where the primary contact has at least one AEO-source tag. Full definition in the CFO's Guide to AEO Budget Defense and the mechanics in the HubSpot-Native AEO Setup. This is the number that survives a board meeting.
Not share of voice — share of pipeline. Of all AEO-influenced pipeline dollars, which prompts did the contacts first-touch on? This is where the 7-prompt rule shows up. If 80% of your AEO-influenced pipeline traces to the same 7–12 prompts, kill the other 488 and reallocate content investment. That's the prompt audit.
Total AEO spend (tool plus content plus headcount allocation) divided by gross margin on AEO-influenced closed-won revenue. Compare to paid-channel CAC payback on the same denominator. If AEO CAC payback beats paid, the answer to "should we reinvest AEO spend in paid?" is no. If AEO CAC payback lags paid, you cut or restructure. Either way, you've replaced a dashboard debate with a CFO-legible answer.
You can keep share of voice as an appendix-level operational metric. It's fine for the person running the program to watch SoV move week over week. It's lethal as a leadership-dashboard KPI.
Partly because SoV is measurable without integrating into a CRM. A monitoring tool that runs prompts against the engines and counts citations can report SoV without ever touching HubSpot. A tool that reports pipeline attribution has to sit inside your CRM and write to opportunities — which is a deeper integration, harder to ship, narrower market. SoV is the easier product to build.
Partly because SoV is the metric the SEO-era team already knew. Ahrefs and Semrush trained a generation of marketers to think in share-of-voice terms. When those teams became "AEO teams," they brought the metric with them. The buyer is familiar with it. The tool makers optimized for what the buyer asked for.
Partly because dashboards that show dollar numbers are much more politically risky for the vendor. A share-of-voice chart going up means your tool is working. A pipeline number going flat means something needs to change — and that conversation gets uncomfortable. Tools that avoid the uncomfortable conversation sell faster in quarter one and churn faster in quarter four.
Whatever the cause, the effect is the same: most tools in market are optimizing for a metric the category has outgrown. The CMOs still buying them are the ones who haven't had their first renewal review yet. After the review, they switch.
We're building Lantern because the replacement KPI needs a tool that can actually compute it. AEO-influenced pipeline, by prompt, with CAC payback — this lives in your CRM. Ours is the layer that writes citation data to HubSpot so the math can run. Pricing: $99/mo or Enterprise. The essay exists because we'd rather be wrong in public about SoV than polite in private.
If you want the technical build, read HubSpot-Native AEO Setup. If you want the budget defense, read The CFO's Guide to AEO Budget Defense. If you want the definition of the metric that's replacing SoV, read the AEO Pipeline Attribution glossary page.
Share of voice measures where you showed up. Pipeline measures whether anyone bought. Pick the second one.
Lantern ships AEO-influenced pipeline dollars into HubSpot — the metric that replaces share of voice. $99/mo or Enterprise. 10 V1 design-partner spots open.
Join Waitlist