The thesis

Media intelligence in Africa is at the same inflection point that financial-services research hit in the 1990s.

For two decades, the industry sold volume — clip counts, AVE numbers, dashboard exports. The buyers tolerated it because there was no alternative. The 2020s are exposing the limits of that model. AI-generated media is polluting the corpus. African languages are growing back into the broadcast stack. Reputation has displaced brand as the C-suite’s frame. Boards have stopped accepting AVE as a credible metric.

What replaces volume metrics is methodology. What replaces self-service portals is analyst service. What replaces "we cover Africa" is country-level credibility, language-by-language depth, and an honest depth-chart of where coverage is full and where it is partner-supported.

That is the thesis behind everything MarketIQ has built since 2005, and what we are betting our next decade on.

Six structural forces shaping the next five years

Each force changes what media intelligence has to be. None of them is a marketing slogan.

01

AI-generated media is now a measurable share of coverage.

Press releases, op-eds, listicles, even mid-tier news content — a non-trivial fraction of what we monitor in 2026 is at least partly machine-generated. The implication for measurement: provenance matters as much as sentiment. The next-generation impact score has to weight authentic human signal differently from synthetic noise.

02

Broadcast voice-to-text is finally good enough to displace human transcription — almost.

Five years ago, automated transcription of African-language broadcast was unusable. Today it is good for English, conversational at best in Afrikaans and Swahili, and still a research project in tonal languages. The arc is clear; the timeline is not. We expect 2027-2028 to be the inflection.

03

Distribution is fragmenting away from the open web.

WhatsApp channels, Telegram groups, private Substacks, paywalled audio. The "monitor the public web" model that defined media intelligence for two decades does not capture where reputation is now formed. The vendors who survive the next five years will have rethought their capture stack.

04

The board has stopped accepting "AVE" as a serious metric — finally.

For fifteen years, the industry has been arguing about the obsolescence of Advertising Value Equivalent. As recently as 2023, half of client briefs we received still mentioned it as the primary metric. By 2025, that number was below 20%. The composite-score era has arrived, behind schedule, but unmistakably.

05

African media markets are diverging, not converging.

The conventional wisdom — that African media is becoming more uniform as global platforms penetrate — is wrong. Local-language broadcast is growing. Regulatory regimes are diverging. The global SaaS platforms that abstract African coverage into a single feed are leaving information on the table.

06

Reputation has overtaken brand as the C-suite’s media question.

Marketing measures brand. The CEO measures reputation. The shift in budget — from CMOs to CCOs — means the questions media intelligence answers have changed. "How are we positioned?" has displaced "How much coverage did we get?" as the brief of the decade.

Specific predictions, with horizons

Time-stamped so you can hold us to them — or watch us be wrong.

+ 12 MONTHS

Sentiment becomes a regulated category.

At least one African regulator will publish guidance on AI-generated sentiment classification in regulated industries (financial services first). Vendors without documented methodology will lose enterprise tenders that previously they would have won on price.

+ 24 MONTHS

The platform-only model becomes a partial offering.

Pure-software media intelligence will reposition as a feeder layer beneath analyst-led services. The "self-service portal as the entire offering" pitch will become increasingly difficult against analyst-led incumbents in markets that value methodology rigour.

+ 36 MONTHS

Provenance scoring enters the standard metric set.

Every credible vendor will have a provenance dimension in their composite score: human-authored versus AI-assisted versus AI-generated. We are working on ours; expect it to launch within the year.

+ 60 MONTHS

African media intelligence becomes a procurement category in its own right.

For multinationals operating in Africa, "African media intelligence" will be a separate procurement decision from global media monitoring — much as Asia-Pacific media intelligence became a category of its own in the 2010s. The vendors who win are the ones already credible on the continent.

Open questions we are working on

Not marketing — actual unresolved methodology questions our team debates in working groups.

Question

How do we adapt the impact score for AI-authored coverage?

Where we are landing

A provenance multiplier. Software-generated coverage gets discounted; human-authored coverage retains full weight. The methodology working group is building the calibration through 2026.

Question

Where does private-channel coverage (WhatsApp, Telegram) fit?

Where we are landing

Where access is lawful and ethical, we will cover. Where it is not, we will not. We will not pretend to monitor private spaces, and we will not build features that depend on consent we cannot verify.

Question

What happens when classification accuracy hits 99% and humans are no longer needed?

Where we are landing

It will not. The 1% error rate is a function of the 99% confidence threshold, not the absolute frequency of mistakes. Boundary cases — sarcasm, code-switching, regional idiom — remain in the human layer. We expect this for the next decade.

Question

Should media-intelligence vendors take editorial positions on what they monitor?

Where we are landing

No. We are the measurement layer, not a participant in public discourse. Our independence is part of the value we sell.

The bet

Methodology, transparency, and analyst-led service — at African coverage depth — is the durable position.

We have watched the platforms come and go. The "AI-only" ones promised to displace us in 2018. The global enterprise platforms promised to absorb us in 2021. Neither happened. The reason is not that we are smarter or larger; it is that the position we hold — credible African coverage, defensible methodology, analyst-led service, transparent pricing — is not the position that volume-and-velocity competitors are trying to occupy.

For the next five years, the fastest-growing problem in media intelligence is going to be quality, not quantity. Boards will keep asking harder questions. Regulators will keep tightening. AI-generated coverage will keep contaminating the corpus. The vendors who win will be the ones who never sold volume in the first place.

Read the full quarterly report.

The 2025 Annual Report goes deeper on each of these forces, with the data behind the predictions. Available on request from sales@marketiq.co.za.