Teladoc Health’s October 28, 2026 results report 0.336 million BetterHelp Paying Users for the three months ended September 30. That is 336,000, but the table’s heading and footnote matter as much as the number.
Keep the monthly window beside the figure
Note 6 defines the measure as the average number of global monthly paying users of BetterHelp therapy and psychiatry services during the applicable period. It includes people paying directly out of pocket and those using insurance coverage. The table separately reports a three-month average and a nine-month average.
Primary source: Teladoc Q3 2026 results, operating metrics table and note 6
The three-month figure therefore should not be relabeled as the number of unique people who paid at any point in the quarter, or as a count taken only on September 30. That is an interpretation of the disclosed averaging window, not a new customer estimate.
Try a small fictional example
Imagine a fictional service with three paying users in its first month, six in its second and nine in its third. Its average monthly count is (3 + 6 + 9) ÷ 3 = 6. Now ask how many distinct people paid during the whole quarter.
In one invented scenario, the same first three people remain, three more join in month two, and another three join in month three. Nine distinct people appear across the quarter. In another, each month has an entirely different group. Eighteen distinct people appear. Both scenarios produce the same monthly average of six.
These invented groups are not BetterHelp data. They demonstrate why monthly averages alone cannot identify overlap between months. Multiplying an average by three does not remove that uncertainty or produce a verified quarterly unique-user count.
Do not infer a payment split from the combined count
Because the definition includes both payment routes, the combined paying-user figure does not by itself reveal how many users belong to each route. Keep any payment-route revenue information in a separate column, with its own units. A dollar amount and a user count answer different questions.
For an earnings-reading worksheet, record the metric name, units, period and definition before writing a headline. Then state the question the figure can answer. If your question is about quarterly unique users, request a matching disclosed measure rather than converting the monthly average into one.
Preserve the source and footnote beside the worksheet. When comparing later reports, check whether the definition and period still match. Leave an unavailable breakdown marked as unavailable. This is a business-metric explanation with fictional arithmetic, not an estimate of individual behavior, treatment results or investment value.
What Does the Average Establish?
