By 
Kenyan mobile subscribers have spent the better part of 2026 complaining that their data bundles are vanishing faster than the numbers on the package suggest they should. Safaricom's most recent public response to the complaint, issued in June, told customers to restrict background data on apps that do not need constant internet access, update apps only on WiFi, reduce YouTube streaming quality, disable automatic media downloads on WhatsApp, and turn off autoplay on social media platforms.
The advice is technically accurate. It is also, to a large share of the subscriber base, an answer to a question nobody asked. Kenyans are not confused about why a phone consumes data in the background. They are asking why a unit of data that cost roughly the same ten years ago now needs to be rationed to survive a single day of ordinary use, at a time when the web itself has become several times heavier than it was when those prices were set.
What actually happened
The frustration has a concrete, recent trigger. In late October 2025, Safaricom quietly cut the allocations on several of its popular data bundles, most notably the "No Expiry" packages that let customers buy data at fixed prices with no time limit. Subscribers who previously received 255MB for KES 51 found themselves getting 102MB for the same price, more than halving the value of the bundle. A KES 100 top-up that used to deliver 200MB effectively became far less generous, and 500MB moved to a KES 250 price point.
Safaricom initially avoided commenting directly, telling customers on social media platform X that it was "aware of the issue affecting the awarding of data bundles" and that a resolution was underway. After sustained public criticism, the company confirmed to local tech publication that a technical issue, not an intentional price change, had caused the cuts, and it restored the original allocations while refunding affected customers the difference in data they had missed out on.
That episode fed directly into the broader complaint building among subscribers: that Safaricom's data economics have started to feel arbitrary and one directional. In January 2026, the telco introduced transaction caps of KES 250 on USSD and KES 500 on the mySafaricom app for non-expiry bundle purchases, again drawing criticism from users who noted the caps meant more frequent, smaller purchases. By June 2026, when a customer publicly questioned why a 2GB bundle for KES 100 had disappeared from his line, Safaricom's answer was that its offers "keep changing periodically and the process is automated," a response that did little to explain why the changes so consistently trend toward less data for the same money.
The technical case Safaricom is not wrong about, and the one it is dodging
Safaricom's underlying technical claim, that modern smartphones and modern apps consume more background data than devices did a decade ago, holds up. Automatic app updates, cloud photo backups, push notification syncing, and operating system telemetry all draw on mobile data even when a phone is sitting idle on a table. Android's Settings menu under Data Usage will show most users a nontrivial chunk of consumption attributed to apps they never opened that day.
But that framing sidesteps the harder question, which is not about background processes. It is about why the price of a gigabyte has not moved in proportion to how much more a gigabyte is now expected to carry. A standard web page a decade ago loaded perhaps 1 to 2MB of data. Modern pages built on JavaScript-heavy frameworks, high resolution imagery, and third party tracking scripts routinely load 5 to 10MB for a single page view. Social platforms that once served static, compressed images now default to autoplaying 1080p video at 60 frames per second, a format that can consume a full gigabyte in twenty to thirty minutes of continuous scrolling on TikTok, Instagram Reels, or YouTube Shorts.
None of that is a fault of the user's habits. It is the direction the entire internet has moved, largely because platforms compete on richer, more immersive content and assume the underlying pipes will keep up. Telecom pricing in Kenya has not kept up in the way that logic would suggest it should. When Safaricom's response to that mismatch is to suggest users throttle their own video quality and disable autoplay, it is effectively asking subscribers to behave like it is still 2016, on a network priced as though it still is.
The market structure behind the pricing
Kenya's mobile market is not short on competition on paper, but it is heavily concentrated in practice. According to the Communications Authority of Kenya's sector statistics for the third quarter of the 2025/2026 financial year, covering January to March 2026, Safaricom controlled 68.9 percent of the country's 84.1 million active mobile subscriptions, with a subscriber base of 57.9 million after adding roughly 5.5 million new lines in a single quarter. In mobile broadband specifically, Safaricom held 62.7 percent of the market, though that figure had eased from 64.3 percent the previous quarter. Airtel, its nearest rival, held 27.6 percent of overall mobile subscriptions with 23.2 million subscribers.
That dominance shows up directly in bundle pricing. A side by side comparison published by Business Daily found that Airtel offers 1GB valid for one hour at KES 15, while Safaricom's equivalent bundle, at 1.2GB for the same validity window, costs KES 20. For 24 hour bundles, KES 20 buys 200MB on Safaricom versus 300MB on Airtel. Independent bundle comparison site PesaMarket, tracking 65 packages across Safaricom, Airtel, Telkom, and Faiba as of March 2026, found Telkom's Mambo 99 bundle (7.3GB for KES 99) and Airtel's Smarta 1500 (90GB for KES 1,500) offering the lowest cost per gigabyte in the market, at roughly KES 14 and KES 17 per GB respectively. Safaricom's comparable 5G 200GB bundle at KES 10,000 works out to about KES 50 per gigabyte.
Despite consistently pricier bundles, Safaricom's market share has not meaningfully eroded. Its strength lies less in data pricing and more in network coverage, particularly in rural counties where competitors have limited infrastructure, and in the lock-in effect of M-Pesa, which held an 89.1 percent share of Kenya's mobile money subscriptions as of March 2026, according to the same CAK data. That combination gives Safaricom limited structural pressure to compete aggressively on data pricing, since the calculation for most subscribers weighing a switch involves far more than the cost of a gigabyte.
Data has become the business
Safaricom's own financial disclosures show why data pricing sits at the centre of its commercial strategy rather than at its periphery. In the half year to September 2025, mobile data revenue overtook voice revenue for the first time in the company's history, closing at KES 44.45 billion, up 18.2 percent from KES 37.6 billion the year before, while voice revenue grew a marginal 0.5 percent to KES 41.09 billion. For the full financial year ended March 31, 2026, Safaricom PLC reported group service revenue of KES 414.1 billion, up 11.1 percent year on year, with group net income growing 67.3 percent to KES 99.7 billion. Mobile data revenue for the year grew 14.4 percent, and the company's own reporting describes it as the largest contributor to connectivity revenue, with Data ARPU (average revenue per user) rising as customers migrate to 4G and 5G devices.
This is the commercial logic that industry watchers point to when explaining why bundle sizes have not scaled with actual consumption needs. As voice and SMS revenue flattened years ago under pressure from WhatsApp and other over the top messaging services, data became the primary growth lever telcos had left. Telcos in Kenya, as elsewhere, pay for international bandwidth capacity in wholesale terms, gigabits per second of throughput, largely denominated in US dollars, then resell that capacity to retail customers in volumetric units, megabytes and gigabytes, bundled with expiry windows. That structure allows operators to extract far more retail value from data than its incremental wholesale cost would suggest, particularly through breakage, the portion of a time-limited bundle that expires unused. Short validity bundles push subscribers into a repurchase cycle that keeps unit margins wide even when actual per-gigabyte transmission costs continue to fall industry-wide.
A regulatory survey referenced by Kenya Insights in January 2026 found that only 18 percent of Safaricom customers report receiving regular, detailed billing information, compared with 44.1 percent of Airtel subscribers and 35 percent of Jamii Telecommunications customers. That opacity compounds the frustration: subscribers cannot easily verify whether their data depleted because of genuinely heavier usage, background processes, or billing inconsistencies, because the operator with the largest market share also provides the least visibility into consumption.
What would actually move the needle
Kenya is not without avenues for correcting this. The Communications Authority of Kenya has statutory power to scrutinise dominant market positions and can, in principle, compel more transparent billing or investigate whether pricing structures constitute an abuse of market power, though it has not to date imposed data pricing controls on Safaricom specifically. Infrastructure competition is also shifting the picture at the margins: Starlink's satellite internet service has begun gaining traction in Kenya's fixed broadband segment, and fibre and fixed wireless providers such as Jamii Telecommunications, Wananchi Group, and Poa Internet Kenya continue to compete for household connectivity outside the mobile network.
None of these forces have yet been strong enough to compress mobile data pricing in the way falling wholesale bandwidth costs would suggest is possible. Until either regulatory intervention changes the incentives or a competitor manages to erode Safaricom's coverage and M-Pesa advantages enough to force genuine price competition on data, the gap between what a gigabyte now needs to do and what a gigabyte costs in Kenya is likely to persist, regardless of how carefully individual subscribers manage their background app settings.
Comments