The short version
- Answer-engine referral volume is small relative to Google for almost every site. Anyone quoting you a universal ratio is guessing.
- The visits that do arrive tend to be better qualified, because the user already read a synthesis before clicking.
- Most of the influence is invisible — acted on later as direct or branded traffic with no referrer.
- Measure your own referrals rather than trusting industry figures. It is a filter in your analytics and takes minutes.
Answer engines that cite sources do send referral traffic, and it is measurable when the user clicks through directly. The question every team asks is how much, and the honest answer is that it varies enormously by audience and topic, and that your own analytics will tell you more than any published figure.
Technical and B2B audiences shifted to assistant-led research earlier and further than general consumer audiences. A developer tools company and a local restaurant will see completely different pictures, and an industry average describes neither.
How the two compare
The last row is a genuine loss for anyone used to query data. You can see that a visit came from an answer engine; you cannot see what the person asked. That removes the single most useful input SEO has ever had, and no tool currently restores it.
How to measure it yourself
- Filter your analytics by referrer for the answer-engine domains you care about. This takes minutes and gives you a real number about your own site.
- Compare conversion rate against Google organic. Higher is common; the sample is usually small, so be careful about reading too much into early numbers.
- Watch direct and branded traffic alongside it, because that is where delayed influence appears.
- Segment by landing page to see which content actually gets cited.
- Track the trend over quarters, not weeks. The volumes are small enough that monthly noise dominates.
What to do with the finding
Once you have your own referral numbers, the useful action is usually not to chase more of them. Answer-engine referrals come from being cited, and being cited comes from being a credible, well-indexed source — which is ordinary SEO plus the extractability habits covered elsewhere on this blog. There is no separate acquisition lever to pull.
What the data is genuinely good for is prioritisation. Segmenting referrals by landing page shows which of your pages assistants actually draw on, and those pages are worth deepening. Cross-check against Search Console for the same URLs — a page that performs in both is your strongest asset, and a page cited by assistants but ranking poorly in search is usually one good technical fix away from doing both.
Set expectations internally before you present any of this. A channel contributing a few hundred visits a month will look trivial next to organic search in any dashboard, and the temptation is either to dismiss it or to overstate its trajectory. Neither helps. Report it as a small, growing, poorly-attributed channel whose main value is influence you cannot fully see, and revisit the sizing each quarter rather than each month.
Small volume is not the same as small value
A hundred visits a month from an answer engine, arriving pre-qualified after reading a synthesis that recommended you, can be worth more than a thousand shallow organic visits. Judge the channel on conversion and revenue rather than on session count — the session count will always look unimpressive.