Buyer-Intent Heat Index
The Buyer-Intent Heat Index tracks where buyer demand is heating up and cooling down across UC, CX and CPaaS, using real Google search behaviour. It shows which tier-one vendors buyers actually search for, and where each one sits in the buying journey. Refreshed weekly.
Each pillar shows a character badge (for example "Vendor-led, category rising" or "Quiet, category flat"), a lead sentence stating the clearest pattern that week, and a tally of how many vendors sit in each buying stage (Awareness through Decision), right now, not a market position.
​
Category terms (broad searches like "AI agents," not brand names) show a heat dot (hot, warm, cool, or cold), based on score and momentum together, never one alone, with the year-on-year change in the note. Vendor bars show each name's search interest as an index against a reference vendor for that pillar (not search volume; the reference always reads 100), and a stage badge marking which search type (comparison, pricing, or support), is strongest for that vendor; darkest badge (Decision) means comparison-shopping language is showing up strongest.
​
The line beside each vendor score is the last 52 weeks of search interest, smoothed. A Rising or Falling tag compares the last 13 weeks with the same weeks a year earlier, and appears only when the move is larger than that of general search terms. "What buyers are also searching" pulls directly from real comparison and alternatives volume.
​
Read the signals together, not the score alone. A high score with a Decision badge and a Rising tag tells a different story than the same score sitting quietly in Awareness.
Buyer-Intent Heat Index: FAQs
It is a weekly read on where buyer demand is heating up or cooling down across unified communications, customer experience and CPaaS, built from real Google search behaviour rather than vendor claims. It shows how often buyers search for each tier-one vendor, at which stage of the buying journey they show up, and how each vendor's search interest has moved over the past year.
Every figure comes from Google Trends search interest, worldwide, over the last 24 months, pulled through the DataForSEO API each week. Nothing is modelled or estimated. The current, unfinished week is left out because Google under-reports it. Google Trends is a sample, so figures can move a few points between pulls.
Each score is relative search interest, not search volume. Google Trends only compares searches fetched together, so every group of searches includes a reference term for its pillar and each score is an index against it. The reference always reads 100, and other terms can score above or below it. In UC the reference is Mitel; in CPaaS it is Twilio, with smaller vendors measured against Sinch; in CX, vendors are measured against Avaya and category terms against the chatbot search. Compare scores within a pillar, not across pillars.
A tag compares the last 13 complete weeks with the same 13 weeks a year earlier, which removes seasonality. It appears only when the change is larger than that of general search terms such as weather, news and maps over the same period, and never for moves under about 30 percent. Terms without a tag are steady.
A tag compares the last 13 complete weeks with the same 13 weeks a year earlier, which removes seasonality. It appears only when the change is larger than that of general search terms such as weather, news and maps over the same period, and never for moves under about 30 percent. Terms without a tag are steady.
Because buyers rarely type the category acronyms. Search demand runs through vendor names and problem language, not the labels the industry uses. For scale, Google Ads shows about 6,600 searches a month for UCaaS and about 673,000 for Twilio (September 2026). On the index the acronyms sit near the floor, which is itself a finding: the category words are close to invisible in real buyer search.
Some vendor names are also everyday words or numbers: Intermedia, Bandwidth and 8x8. Most searches for them have nothing to do with the vendor, so their scores would mislead. They will be added back when a company-specific measure reads cleanly.