How to read a market forecast without being misled by one

Almost every vendor presentation in this category opens with a market-size figure. Here is how to read one.

As published:

• Juniper Research forecast CPaaS revenue growing 13% in 2026 to about US$34 billion, up from about US$30 billion in 2025.
• Gartner forecast CPaaS spending of US$15.7 billion in 2025, rising to US$27.4 billion by 2029 — about 14% a year. (Its 2026 update revised 2025 down to about US$14.9 billion; see what the analysts disagree about.)
• Mordor Intelligence sizes the market at US$21.27 billion in 2026, reaching US$41.05 billion by 2031 — also about 14% a year.

These are serious firms. The spread is not sloppiness. Notice the other thing: they disagree on size by billions of dollars, yet agree on growth of roughly 13–14% a year. That pattern is the most useful thing in the slide.

• Pass-through fees: the part of a message's price that the platform collects and hands to a mobile operator or to Meta. The platform never keeps it.
• CAGR (compound annual growth rate): the steady yearly growth rate that would take the market from the start figure to the end figure.
• Base year: the year the forecast starts from. If that figure is itself an estimate, every later year inherits its error.

Pass-through fees. Count operator fees and the market looks very large; count only what platforms keep, and it is much smaller. Both are defensible; they answer different questions.

Where the boundary sits. Is a contact-centre suite CPaaS? Email delivery? A video SDK? Each firm draws the line differently, and each line is a different market.

Spend versus revenue. What buyers spend and what vendors book diverge, because one includes things the other does not.

1. What is inside it? If a presentation cannot say whether pass-through fees are included, the figure is decoration.
2. What is the unit? Revenue, spend, messages, seats and "market opportunity" are five different things. Slides switch between them without saying so.
3. Is the base year actual or estimated? A five-year projection from an estimated base is an opinion with a chart attached.
4. Which population? A global total says little about whether your segment, in your region, is growing.
5. Is the growth rate for the whole, or for its fastest part? A common, usually innocent, sleight of hand: measuring the fastest-growing segment and applying it to the whole category.
6. What is it being used to argue? A growth number in a sales deck is there to make a purchase feel inevitable. That is not a reason to disbelieve it — just to notice the work it is doing.

Direction and rough scale. When firms with different methods agree on direction, that is a genuine signal. Forecasts can also tell you whether a category is consolidating (fewer, bigger players) or fragmenting.

What they cannot tell you is whether you should buy, build or do nothing. None of them modelled your volumes, your countries or your costs. Use forecasts to understand the weather, not to decide what to wear.

The figure you actually need is your volume, price and cost — and you can work it out from things you already know.

An illustrative example for an online shop:

Add other triggers — appointment reminders, renewals, support conversations — the same way. That number is worth more than any published total because it is about you. Then use the published forecasts for a sanity check on direction.

• The state of business messaging in 2026 — the next step in this track.
• What analysts disagree about — the deeper look at methods.
• What CPaaS actually is — why pass-through fees are so large.