African Technical Needs Survey: Analysis & Findings

When we started planning the WikiIndaba Hackathon 2026, we wanted to think about how to make the event as useful as possible for participants. Beyond the usual hackathon experience, we wanted people to get hands-on support, work together and, most importantly, spend time improving tools that African Wikimedia communities actually need. We didn’t want to make assumptions about what those needs were, so we decided to ask the community. Wiki Mentor Africa (WMA), on behalf of the hackathon planning committee, put together a Technical Needs Survey to find out which tools people currently use, what challenges they face and what they’d like to see improved.

We received 93 responses from 12 countries between 8 September and 8 October. A few things stood out from the responses. For 70 respondents, internet connectivity and device limitations affect the tools they can use. Another 45 said they had avoided using a tool because it, or its documentation, wasn’t available in their language.When asked about the biggest technical gaps, low-connectivity and offline editing support (44 responses) and local language support (43) came up most often, followed by Wikidata and data-quality tools (30). Respondents also mentioned everyday tasks they’d like to make easier, particularly translation, finding and adding references, uploading and categorising files on Commons, and tracking events and contributors. There was also strong interest in taking part in the hackathon: 87 people said they wanted to participate, while 83 said they would be willing to give their time to improving tools for the region.

These responses give us a better idea of where the community needs support and will help us decide which existing tools to focus on at the hackathon.

Who responded

Respondents listed 39 different groups, with participants able to name more than one. Nigeria and the DRC made up the largest share of responses, with 62 of the 93 responses (67%) coming from the two countries. This means we have fewer responses from other countries. The “Others” category in the chart below covers six responses: one each from Rwanda, Burundi, Senegal, Tunisia, Tanzania and Togo.

Time in the movement: Most respondents have been involved in the Wikimedia movement for 1–3 years (38), followed by those with 3–5 years of experience (21). There were also 17 respondents who had been involved for less than a year and 17 for more than five years. Roles (participants could pick several): 84 editors or content contributors, 27 event or community organizers, 13 technical contributors, 6 affiliate leaders or board members. Home project: Wikipedia was the most common home project (43), followed by Wikidata (28) and Wikimedia Commons (11). Wikiquote had 7 respondents, while Wiktionary, Wikisource, Meta and MediaWiki had one each.

What tools do people already use

Survey question:  Do you currently use any tools, gadgets, or bots in your Wikimedia work? 

Survey question:  List the tools, gadgets, or bots you use in your Wikimedia work. 

Survey question:  Where do you usually discover tools relevant to your work? 

Of the 93 respondents, 47 said they currently use tools, gadgets or bots in their Wikimedia work, while 46 said they don’t. 35 people went on to list the tools they use.

The tools mentioned most often were Outreach Dashboard and Content Translation (6 mentions each), followed by HotCat (5) and the event registration tool (4). XTools, QuickStatements, Spell4Wiki and Wikidata Query Service/SPARQL were each mentioned three times. Other tools included Broomstick and ISA Tool (2 mentions each), while several others were mentioned once. Seven responses also mentioned devices such as phones and laptops, and two mentioned AI assistants.

We also asked where people usually discover tools. Among the 46 who answered this question, the most common answer was through other editors or word of mouth (23), followed by social media (17) and Phabricator (15). ToolHub was mentioned by 10 respondents, while 9 said they don’t actively look for tools.

This suggests that people still rely quite a bit on other contributors and informal channels to discover tools. ToolHub is reaching some users, but there may be room to make it more visible and easier to discover.

The biggest technical gaps

Survey question:  Which of these areas feels like the biggest technical gap in your own work? 

The top two options were each picked by about half of respondents (44 and 43 of 93) and the third by about a third (30). Counts add up to more than 93 because participants could pick several options.

Connectivity and mobile constraints

70 of 93 respondents (75%) said connectivity or device constraints limit which tools they can use. 35 of those 70 also picked low-connectivity/offline editing as a biggest gap. In the free-text comments, 9 of 52 mention mobile and 7 mention offline use, low bandwidth or a limited connection.

These are some of the accounts in respondents’ own words (the first two translated from French using Ai translation):

• DRC: “for each photo, we must manually enter on Commons: the description in French, the categories (e.g. Lubumbashi, Katanga, Markets in DRC), and the GPS coordinates. With an unstable connection and from mobile, this can take 8 to 10 minutes per image.” The respondent proposes a mobile feature or small bot that suggests a basic French description and relevant DRC categories at upload, and says it would take them “from 10 photos per session to more than 50“.
• Goma, DRC: “a large part of the contributors, about 80%, contribute only on a mobile phone with a limited connection. The current tools are not optimized for this.“
• Uganda: “Most of us in Uganda edit only from phones with unstable data. We need lightweight tools that work on whatsapp/low bandwidth, batch photo cleaning/upload, and support for Runyankore and other languages.“

These are individual accounts. The survey did not measure what share of each community edits from a phone.

Language and localization

Survey question:  Have you ever avoided using a tool because it wasn’t available in your language or the documentation wasn’t? 

45 of 93 respondents (48%) said they have avoided a tool because it, or its documentation, was not available in their language, and 43 picked local language support as the biggest gap. In the manual-task answers, translation was the most common theme (9 of 60). In the comments, 10 of 52 raise local languages and 6 raise translation tools. Languages named include Igbo, Igala, Hausa, Swahili, Lingala, Runyankore and French.

In respondents’ own words:

• Nigeria: “A tool that can be used to translate several articles on igbo Wikipedia will be appreciated“
• DRC (translated from French): “The article translation tool (Content Translation) works poorly for translating into French and even less into Swahili and Lingala, which are spoken by millions of participants in DRC.“
• Tunisia: “Wikipedia’s translation tool needs further updating, especially regarding the search for topics the user wants to translate.“
• Uganda: “Some tools/gadgets/extensions are not used because they are not configured on the small wikis and getting help is challenging“

Content Translation is among the most-named tools in use (6), and two respondents (DRC, Tunisia) criticize the Wikipedia article translation tool. My understanding from the responses is that for translation, the request looks like better support in existing tooling rather than a new tool.

What participants said in their own words

Survey question:  What’s one manual task you do regularly that you think could be improved with a tool, bot, or software feature?

60 of 93 answered the question about a manual task they would like a tool to improve. 7 of those were “none” or unclear (for example “Néant” or “I have no idea“). I grouped the other 53 by theme, one theme per answer. 

Theme	Answers
Translation	9
References, citations, finding sources	7
Commons upload, media descriptions and categorization	7
Event and contributor tracking, dashboards	6
Wikipedia categories and article structure	3
Spelling, grammar, language quality	3
Wikidata lexemes and queries	3
Offline editing	1
Other one-off tasks (for example Wikiquote visual editing, social media posting)	14

The DRC photo-upload account in the connectivity section is one of the 7 Commons answers. The Uganda account came from a final comment and is not counted in the table.

52 participants left a final comment, and 26 of those were “no” or “nothing” style replies. 9 of the comments ask for training, onboarding help, documentation, or beginner-friendly tools. One respondent wrote: “There is currently a significant gap for new contributors regarding how to utilize these technical tools. To address this, we need to organize more dedicated training sessions and hands-on workshops…” A respondent in Nigeria wrote: “please my community needs a tool that helps pinpoint errors in references so as to avoid members from being blocked”. A respondent in DRC wrote: “Hackathons could also be connected to longer-term technical support so that useful prototypes can evolve into sustainable tools used by communities across Africa.“

Willingness to contribute

Survey question:  If you have technical skills, would you be willing to dedicate time to help develop or improve tools for the region?

Survey question:  Are you interested in joining our upcoming hackathons (WikiIndaba Hackathon and/or WMA End of Year Hack Challenge 2026)?

87 responded with a yes, 4 maybe, 2 no. Willingness to give time to develop or improve regional tools: 83 yes, 9 maybe, 1 no. That question began “If you have technical skills”, and the survey did not ask what skills participants have. 13 respondents described themselves as technical contributors (11 yes, 2 maybe). Among the other 80, 72 said yes, 7 maybe and 1 no, so the yes answers cannot be read as a count of available developers.

Recommended focus areas for the hackathon

From my analysis, these are my suggestions for the hackathon. They are not survey results. Each comes with the survey evidence behind it, ordered by the closed-question counts. The PTAC draft recommendation advises finding and improving existing tools before building new ones, so each item is framed as improving something that already exists.

1. Mobile and low-bandwidth use of existing tools. 70 said connectivity or device limits restrict their tools, and 44 picked offline editing as a top gap. The most detailed accounts are about Commons uploading on mobile in DRC and Uganda (7 manual-task answers are about Commons uploads and categorization).
2. Translation tooling for African languages. 45 avoided a tool over language or documentation, 43 picked local language support as a top gap, translation was the largest manual-task theme (9), and 6 respondents named Content Translation as a tool they use.
3. Wikidata and data-quality tools. The third most-picked gap (30). Manual-task answers include adding Igbo lexeme pronunciations (2) and a request for PetScan to show several languages side by side (1).
4. Event and contributor tracking. 18 picked event/campaign organizing tools as a top gap and 6 manual-task answers describe counting contributions by hand. Outreach Dashboard is already the most-named tool (6), so the first thing to find out is whether the gap is the tool itself or using it on mobile and low bandwidth.
5. References and citations help. 7 manual-task answers, plus the request from Nigeria for a tool that flags reference errors so members are not blocked.

22 participants picked onboarding and training tools as a gap and 9 comments ask for training. That is a training need as much as a tooling need, so it may fit the preparation sessions better than the build itself.

The survey shows needs. It does not show which existing tools can be extended or who maintains them, and that needs checking before the we (the Wikiindaba planning committee) commit. 

Notes on method and limitations

• Source: the Google Form export of 93 responses, timestamped 8 September to 8 October 2026. Two responses arrived after the 7 October close and are included.
• Respondents chose to take part. The survey does not show how many participants saw it, so there is no response rate.
• Nigeria (44) and DRC (18) make up 67% of responses and six countries have one response each, so differences between countries are not reliable but the main findings still hold. Without  Nigeria, low-connectivity was picked by 53% and local language support by 41%, and Wikidata/data-quality tools came third at 37%. When I also removed DRC, only 31 respondents were left, and language support and Wikidata tools tied for second at 42%. 
• The tool-discovery question was answered only by tool users (46 of 47).
• Theme groupings of free-text answers are my own manual coding. French quotes are my translations using AI translation tools. All other quotes are verbatim.