CHW Connect: Usage Analytics
CHW Connect: Usage Analytics
Context
CHW Connect is a digital platform supporting community health workers (CHWs) to register clients, log visits, track child development, and run breastfeeding clubs. It operates within Grow Great Champion’s CHW programme across four provinces in South Africa, two of which ran time-limited pilots that have since concluded. This analysis covers all registered CHWs from launch through May 2026.
Platform scale
CHW Connect has operated across four provinces. Gauteng and Limpopo ran time-limited pilots which have since concluded. Mpumalanga and Western Cape are the two active provinces where the programme is ongoing.
341
Registered CHWs (all provinces, all time)
264
CHWs in live provinces (Mpumalanga & Western Cape)
58
Active in last 30 days
9,118
Child clients registered
3,224
Pregnant clients registered
81,607
Total visits completed
The 341 total includes CHWs from the concluded Gauteng and Limpopo pilots. For active programme analysis, the relevant base is the 264 CHWs in Mpumalanga and Western Cape, of whom 58 (22%) were active in the last 30 days. The visit volume, 81,607 across all provinces and all time, reflects the cumulative work of a small consistently active group.
Part 1: Engagement by province
The chart below shows engagement status for CHWs in Mpumalanga and Western Cape.
Reading the chart
The two live provinces are Mpumalanga and Western Cape. Of 152 Mpumalanga CHWs, 33 were active in the last 30 days (22%). Of 112 Western Cape CHWs, 25 were active (22%).
Active use is concentrated in a small group
The majority of registered CHWs are churned or at risk. 22% active in both Mpumalanga and Western Cape suggests a stable core of consistent users rather than broad adoption across the workforce. The visit volume data suggests this core is highly active.
Part 2: Visit consistency
Reading the charts
Visit volumes have varied significantly over time, with Mpumalanga accounting for the majority throughout. A notable spike in October 2024, reaching over 200 average visits per active CHW, likely reflects retrospective data entry rather than visits completed in that month. The causes of the broader volume patterns are not fully understood from platform data alone and will be explored with the Grow Great programme team.
Active CHWs in Mpumalanga are logging significantly more visits
Mpumalanga CHWs who are active consistently average 50-80 visits per month, with the exception of the October 2024 spike. Western Cape active CHWs average 5-15 visits per month throughout the same period. Whether this reflects different programme intensity, different caseload sizes, or different data entry practices is worth exploring with the programme team.
The active core is using the tool regularly
Despite low overall activation rates, CHWs who are active show consistent month-on-month visit logging rather than occasional use. This suggests the platform fits into the workflow of CHWs who are using it, even if broader adoption across the workforce has not materialised.
Part 3: Client load
How many clients are CHWs managing, and how is that distributed across the workforce?
Reading the chart
In Mpumalanga, average child clients per CHW is 40.5 but the median is 17. A small number of CHWs are carrying very large caseloads, up to 223 child clients for one CHW, while the majority have far fewer. Western Cape shows a similar but less extreme pattern (average 16.7, median 13). For pregnant clients both provinces show a smaller gap between average and median, suggesting more even distribution of maternity caseloads.
Caseloads are unevenly distributed
The average vs median gap in Mpumalanga confirms the super-user pattern: a small number of CHWs are driving a disproportionate share of client registrations. Whether this reflects programme design (some CHWs assigned larger areas) or differential engagement (some CHWs are more active than others) is worth investigating at the programme level.
Part 4: Feature adoption
Reading the chart
Registering clients and logging visits are the most widely adopted actions on the platform. In Mpumalanga, 100% of CHWs have registered at least one child client and 90.1% have logged a child visit. Western Cape shows similar patterns at 92% and 72.3% respectively. Pregnant client registration and visit logging follow a similar pattern in both provinces.
Breastfeeding clubs sit lower in the adoption picture: 38.8% of Mpumalanga CHWs and 35.7% of Western Cape CHWs have run at least one club. In absolute terms Mpumalanga has logged 1,421 clubs and Western Cape 238, with an average of around 5 attendees per club. Calendar usage sits at a similar level to breastfeeding clubs (36.8% in Mpumalanga and 35.7% in Western Cape) suggesting a consistent subset of CHWs are using the platform beyond the core features.
Core workflow features have strong adoption
Client registration and visit logging are embedded across the active CHW workforce in both live provinces. Among CHWs who are active, the platform is being used for its core intended purpose. The high client registration rate in Mpumalanga (100%) suggests that registering clients on the platform has become a standard part of the CHW workflow there.
Breastfeeding clubs are a meaningful secondary feature
While breastfeeding club adoption is lower than core workflow features, the absolute numbers are significant. 1,421 clubs logged in Mpumalanga represents sustained use of a feature that helps CHWs organise and record group-based activities. The similar adoption rates in both provinces (38.8% vs 35.7%) suggest this feature resonates across different programme contexts.
Part 5: Team Lead activity
Reading the chart
Team Lead engagement is low across all provinces. Mpumalanga has 9 Team Leads with an average of 438 days since last seen, over a year. Western Cape is more active at 133 days average, with 3 of 4 Team Leads showing some activity in the last 90 days. The Team Lead layer appears to be largely dormant as a platform feature.
Team Leads are not using the platform
The supervisory layer intended to support CHWs is not active on the platform in most provinces. Whether this reflects a deliberate decision to use other supervision channels, or simply that the platform does not yet offer enough value for Team Leads to log in regularly, is worth understanding before designing a supervision feature into a new build.
Implications
1. Active CHWs have integrated the platform into their workflow
Among CHWs who are active, the data shows consistent and broad feature use. In Mpumalanga, 100% of CHWs have registered at least one child client and 90.1% have logged a child visit. Western Cape shows similar patterns. The challenge is sustained use across the wider workforce.
2. A small group is driving most of the activity
The gap between average and median client loads in Mpumalanga (40.5 vs 17 child clients) confirms that a small number of CHWs are carrying a disproportionate share of registrations and visit volume. 81,607 total visits from a workforce where only 22% were active in the last 30 days points to a concentrated pattern of use. Whether this reflects programme design or differential engagement is worth understanding before drawing conclusions about overall platform adoption.
3. Secondary features show consistent minority adoption
Breastfeeding clubs and calendar usage both sit at around 35-39% in live provinces, lower than core workflow features but consistent across provinces. These features are used by a meaningful subset of CHWs. The similar adoption rates across two different provinces suggests these features fit naturally into CHW workflows where they are used.
4. The Mpumalanga and Western Cape visit gap needs explanation
Active CHWs in Mpumalanga average 50-80 visits per month while Western Cape active CHWs average 5-15. Both provinces have similar CHW numbers and similar feature adoption rates. The cause of this visit volume gap is not visible in the platform data and will be explored with the Grow Great programme team.
Methodology note
All data is aggregate and anonymised. No individual CHW, client, or child data is reported. Analysis covers CHW Connect from launch through May 2026. CHWs are deduplicated by phone number. Active is defined as having logged in within the specified time window. Gauteng and Limpopo data reflects pilot periods only; both pilots have ended.
Analysis by Kim Tichmann, Data & Product Strategy Lead, ECD Connect, DGMT.
CHW Connect is a product of the ECD Connect project at DGMT in partnership with Grow Great. All views are those of the author.