During the interviews I conducted in Zimbabwe last October, many participants asked me: “Why pick Harare as a case study? Why not Nairobi or Lagos?” Indeed, Harare’s “ecosystem” is nascent, and the few tech startups that exist are hampered by fundamental market and skill constraints, as discussed in my last post.
At the same time, Harare has been marked by a uniquely fast emergence of tech innovation hubs. Starting in late 2013, within only about a year’s time, three hubs have been launched: Hypercube, Muzinda Hub, and Area 46. Pitch Nights (run by Emerging Ideas) and the ZOL Startup Challenge are other initiatives that, without a constant physical presence, don’t qualify as hubs by most definitions, but these event- and competition-based initiatives still function as important support channels for tech entrepreneurs, thereby complementing hubs.
Given the unique speed of this development and my least-similar case study research design, I felt that Harare would be great to examine if and how hubs work when implemented in challenging environments but with a lot of upfront goodwill and drive. This post will examine the city’s hubs more closely, before the last one on Harare will discuss a specific salient discussion point for the local support landscape: the “NGO money” debate.
Like earlier posts, this post serves to accurately and comprehensively describe, document, and take stock of networking patterns in a complex system, so be warned that it is on the long side.
Hypercube is probably the most prominent hub of Harare. It draws very directly from typical and popular operational elements of hubs. Its governance structure bets on a purposely diverse implementation team (locals and expats; men and women; backgrounds in coding, social & digital media, crowdfunding, IT) with international experience that emphasizes grassroots ideals and community. The funding comes mainly from donors (Hivos, the US Embassy to Zimbabwe, and Indigo Trust), while 263 Chat functions as an event collaborator and Comp Trouble Zimbabwe provides in-kind contributions in the form of cheaper rates for tech support (Hypercube team members run or are part of both companies).
The hub’s mission is to support and integrate the tech entrepreneurship ecosystem, while emphasizing social development causes such as the inclusion of women in tech. Participants and other’s blog posts stressed how important self-sustainability and -reliance are for the hub: donor contributions were supposed to only function as an initial subsidy, and funding agencies were not meant to interfere with operational decisions. At the same time, the reputation and contact networks of funders were envisioned to enhance the hub’s international brand and help attract big-shot speakers. In sum, I got the impression that the Hypercube setup was more or less according to the blueprint of donor-funded, grassroots-driven, (social) entrepreneurship-oriented hubs in developing country contexts, which has made it a very interesting use case for many hub assumptions and ideals but also a target for criticism.
It’s still early days though, and Hypercube has functioned in test mode for its first year of implementation, making it difficult to speak to its effectiveness. The space was opened to everyone without a fee and the hub has mostly focused on events, hosting international or locally known speakers (mostly charging participants a small fee). Unsurprisingly, in an activity-starved environment like Harare, interview participants reported that these events were the hub’s most significant value addition; in particular the 2013 StartupBus (now called Ampion Venture Bus), Zimbabwe’s first Startup Weekend, and Techwomen Zimbabwe events. The co-working space has also become an essential asset for some fledgling entrepreneurs: not only do they save rent and bandwidth cost, they can also invite clients to the hub for meetings. Apart from a handful of regulars, most techies and entrepreneurs (especially those just starting out) have passed through the space irregularly, attending an animation course, networking, or just hanging out on maybe one or two days of any given week.
Throughout, the Hypercube team has been clear that things would have to change to generate a more direct contribution to startup creation (a key target) and to make money for the hub. Important findings from the test year were that there is interest and demand for a tech-focused hub, while the state of the local scene wouldn’t make traditional incubation, investment, and sustainability models viable.
Hence, following the hub’s re-launch after renovations, the leadership intends to implement a kind of community-driven incubation model that is meant to keep cost low while maximizing support value for entrepreneurs. Use of the space would be limited to paying individuals, who have the option to either co-work independently or form a team and join a startup building process. In regular intervals based on milestones, teams would receive feedback and validation from mentors and other community members. The upstairs area of the hub would function as office space for more settled entrepreneurs, who would also function as informal and ad hoc mentors. Conceptually this sounds like a great approach to me, but only time will tell if complicated innovation networking dynamics (see below) and the daunting systemic market constraints will allow the program to work.
Also Muzinda Hub (formerly Muzinda Umuzi Hub) has undergone a test phase and will have a re-launch in 2015. The hub is funded by Higher Life Foundation, which itself is backed by Econet (Zimbabwe’s ubiquitous home-grown mobile operator) and run by Tsitsi Masiyiwa, the Econet founder’s wife (both are national celebrities). The hub’s board and implementation team is composed of Zimbabweans with backgrounds in business support, coding, entrepreneurship, and social development.
Initially, in late 2013, Muzinda ran two social change-oriented hackathons, which were meant to spark attention and surface talent for an incubation-focused hub model. The hub occupied a house in Strathaven (a residential but fairly central Harare neighborhood), which at full capacity could have hosted maybe a dozen startup teams. However, the hackathons did not resonate as well as expected. It also turned out to be difficult to support teams in depth while helping a large number of techies and innovators. Two startups, Mazwi and Zimall, were repeatedly mentioned as the hub’s primary successes, while other participants told me that Muzinda’s value addition to these companies was more limited and so they shouldn’t be seen as a direct result of Muzinda’s activity.
Ultimately, the challenges led the hub to “pivot” and focus squarely on a “digital skills” agenda and less on startup incubation. Higher Life Foundation made a larger space available in Vainona, a neighborhood in the Northeast of Harare, not too far from the University of Zimbabwe. The hub recently engaged Ubuntu Equity, a South African / Zimbabwean business development consultancy, to conceptualize the new program. The idea is to educate hundreds of university graduates on applied coding and IT skills, while inculcating an entrepreneurship mindset and basic startup skills. The new program is meant to prepare participants to become employees in the tech industry or to start companies on their own. As with Hypercube’s shift, I find this an interesting model, but the program will hinge on sufficient demand and talent of participants. The re-launch was slated for early 2015, so few of my interview participants were aware of the details.
In the bigger scheme of things, it was interesting that Muzinda moved from providing a few startups with intense and customized support to a more educational and standardized support approach for a larger number of individuals. I found that this shift was due, first, to the identified skills gap, which was seen by many as a crucial bottleneck that makes any narrower startup and incubation support measures less effective. Second, organizations like Higher Life Foundation were seen to be geared towards “numbers”, meaning that they are inclined to support only those projects where large-scale effectiveness can easily be communicated by showing high output figures, like the number of course participants or graduates. Philanthropic organizations also compare themselves to other large funders, such as Rockefeller Foundation, and within this frame of reference, quantitative outcome targets like “5 startups created per year” appear unimpressive. (For instance, Rockefeller talks about “Impact[ing] 1 million lives in six countries in Africa” as a goal for its Digital Jobs initiative.)
I got conflicting information about Econet’s role in this mix. Some said, Muzinda is a separate, well-intended and relevant Higher Life Foundation initiative, where goals are related to social impact and not to Econet’s bottom line. Others speculated that Econet had a longer term business interest in supporting talented knowledge workers and entrepreneurs through the hub, but added that this was fair game and ultimately a win-win for Econet and Zimbabwe. The skeptics outlined how Econet was out of touch with grassroots tech entrepreneurs, and that there was a risk that Muzinda becomes an innovation procurement channel of sorts, where entrepreneurs’ projects are absorbed by the tech giant without being nurtured as independent companies. Here again, details in the hub’s implementation will determine which scenario holds more truth.
Even if it doesn’t call itself that, Area 46 can count as Harare’s third tech innovation hub. So far, it has been functioning as a privately run rent-based model, where a handful of Harare’s revenue-generating tech startups are paying to occupy rooms in the hub founder’s private home in Mt Pleasant (a residential neighborhood close to the University of Zimbabwe). Techzim, Esaja, and EatOut—all more established startups for Zim standards—are notable residents. The companies share fiber-to-the-home Internet access, which was mentioned as a difference maker.
I have not heard a single negative or critical voice about Area 46 and its approach. Self-run and -funded entrepreneurial co-working and peer mentorship appears to be the least equivocal way to go for hubs. Yet, also the typical limitations of this model were mentioned: By design, the startup entrepreneurs that can and want to afford to be there form a bit of an exclusive, like-minded, tight group, and so the model doesn’t work for those just finding out about tech entrepreneurship.
Moving forward, the crux appears to be how Area 46 can scale beyond rent-based co-working for a few startups, providing more comprehensive entrepreneurship support services for a greater number of individuals and teams. It turned out that the hub’s current and potential clients are interested in mentorship and business development support, but such services are costly and tough to arrange and procure in Harare. As a result, plans to do more have so far largely remained that: plans.
The most notable exception is a recent collaboration with the 2014 edition of the ZOL Startup Challenge. On a case-by-case basis, finalists would receive privileged access to space at Area 46. This was seen as a potential donor-independent stepping stone for Area 46, but once again, only the next year and beyond will show the effectiveness and scalability of this approach.
Pitch Nights, Emerging Ideas, ZOL Startup Challenge, and TechWomen Zimbabwe
Aside from hubs, I found events and competitions to be important support channels for Harare’s tech entrepreneurs. Pitch Nights is a monthly event where entrepreneurs that are just starting out present their business, trying to get feedback and exposure. Pitch Nights and also The Basement (where the events are hosted) are projects of the Emerging Ideas virtual accelerator. Emerging Ideas has not yet funded and directly supported an ICT/mobile app/content/software startup, but tech entrepreneurs have still had a strong representation on Pitch Nights.
Pitch Nights (similar to Hypercube events) have been extremely popular and all participants commended the organizers on their efforts. The initiative was seen as a game changer for entrepreneurship, as a unique opportunity for networking, and in many respects the first outlet in Zimbabwe to promote entrepreneurship and its “global movement” to a broader audience. After initial questions about the quality of pitches, the organizers have invested more time to prepare presenters, which has recently amounted to the provision of a mini coaching program. Pitching at one of the sessions has become a rite of passage for any Harare-based entrepreneurial team that wants to call itself a startup. At the same time, Pitch Nights are volunteer-run and networking- and awareness-oriented, so interview participants saw the initiative as complementary to more resource-intensive and in-depth support methods.
The ZOL Startup Challenge
is the longest running was probably the first proper tech startup support initiative ( it evolved out of the if we’re ignoring the now-defunct Jumpstart). Founded together with the long-time CEO and owner of ZOL (an early local ISP upstart), it is run by Techzim. This means that the Challenge has a close affiliation with the popular tech blog and Area 46, where the hub has functioned as a host space for challenge-related events during the most recent edition. The Challenge has undergone a learning process similar to other innovation competitions. Emphasizing relatively large pots of prize money in earlier years, the Challenge now has a strong coaching element, providing feedback to pitches of semi-finalists and offering multiple workshop sessions on startup topics for finalists.
This shift was very much in line with comments I collected from former and current participants of the ZOL Startup Challenge: prize money functioned as an important motivator to participate, but the substantive value creation and learning actually happened through peer-mentoring, feedback at pitches, and business connections generated through the participation. Interestingly, it was pointed out that exposure to new non-technical customers was more limited, given that Techzim’s audience is focused on… well, “tech.” As with Higher Life Foundation’s involvement with Muzinda and donor organizations backing Hypercube, some participants criticized that ZOL would “just” be interested in enhancing its own brand, while the startups’ actual market traction would ultimately not be the sponsor’s primary concern, limiting the effectiveness of the whole initiative. While I couldn’t corroborate or falsify such claims, it remained that the popularity of the Challenge has been increasing quickly, with the number of applications multiplying every year.
Finally, Techwomen Zimbabwe is a noteworthy entrepreneurship support initiative. In collaboration with Hypercube, the bootstrapped group of female activists had run the Zimbabwe leg of the Technovation Challenge, resulting in a handful of girl app developer teams, some of which later continued their work under Astro Mobile’s guidance. The group also runs workshops and tech awareness outreach in high schools across Zimbabwe. I found it striking how closely the program resembled other girls-in-IT initiatives across Africa, such as Akirachix in Kenya.
The most interesting dynamic from an innovation network point of view was probably how the initiative created a win-win with Hypercube: the Techwomen’s grassroots work and energetic pursuit of their mission congealed a story that attracted a lot of attention also for the hub. In return, the group was able to use Hypercube’s professional event infrastructure as well as the co-working space for gatherings of the organizers and supported girls.
Overall Ecosystem Trends: Networking, Experimentation, and Ambiguity
It was a clear finding that participants in Harare, much more so than in Kigali, used and believed in the “ecosystem” idea. Despite different assessments of priorities, hubs were keen to network, divide up roles, and coordinate, often including with government and development organizations. Hubs had also just formed the Catalyst group, a sort of national association of hubs, which is an idea that I have not encountered in any other country.
Yet, even after 40 interviews, I wasn’t sure why so many hubs and so much activity had been started in such a short time. An evolutionary and optimistic theory I heard was that the development of entrepreneurial ecosystems follows an exponential growth curve, and Harare has now entered the steep upward slope. A more subjectivist opinion held that it was just a matter of time before Zimbabweans would jump on the African hub bandwagon, and a race to be at the cusp of the movement happened once it became clear that the first local hub was in the works; hence the quick emergence of hubs wouldn’t necessarily mean that they are needed or effective.
No matter what theory on ecosystems and hubs one believes, a key question was what a Harare hub could be other than “just a space.” All hubs aim to enable co-working and co-creation in some form, but the interviews clearly revealed that the magic lies in the given meaning of “co-“. This boils down to questions that all networking interventions face: How can relationships be established and maintained? How strong or weak do these ties have to be to lead to innovation? Relationships with whom are useful? How much networking is needed, and how much amounts to a distraction from business building? How does one create “energy” and motivation in collaborative settings? How can the value of (newly created) relationships be measured? No hub in Harare has figured out a winning formula, but these questions have persistently been relevant for all of them.
Hence, the networking and collaboration benefit of Harare’s hubs that I found was ambiguous. Sometimes informal interactions like those stimulated at Hypercube events and Pitch Nights were highly useful (e.g., leading to startup formation and learning), but entrepreneurs tended to emphasize that such networking cannot always help them overcome more fundamental challenges, such as lacking market opportunities and skills. Channels to improve networks with customers and (non-technical) partners were particularly sparse, and hub-facilitated networking that was limited to other aspiring but inexperienced techies and entrepreneurs was described as nice-to-have but not essential.
More broadly, hubs were challenged by high levels of uncertainty and complexity. It seemed to me that the many reports, ecosystem conceptualizations, indicators, guidelines, and hub blueprints that are out there were of very little help for the people that are faced with the task to design a real hub organization for a real, highly complex setting with real, sometimes insurmountable challenges in the economic environment. In the absence of success stories and “obvious” approaches, hub managers resorted to experimentation.
While few would argue that experimentation isn’t the right way to go in complex innovation (eco)systems, it remained hard to see how even experimentation (incl. potentially high opportunity cost) could prove the effectiveness of any given model or deliver clear learning outcomes and implementation guidelines. Randomized control trials and other comparative methods (as the few widely accepted quantitative impact evaluation techniques) are hard or impossible to fund and implement for hubs, especially for those starting out. The predominantly informal, qualitative assessments are often affected by a given hub leadership’s subjective perspective that is bound to be influenced by external incentives and the institutional environments that it is operating in, plus intrinsic and often pre-conceived motivations and beliefs about entrepreneurship support. In Harare, this was quite evident in the sheer diversity of intervention models and rationales described above.
In my opinion, in Harare, this has led to somewhat of an overreaction, based on a misunderstanding about rigor and accountability: Self-sustainability (i.e., ultimate donor independence) and quantitative impact measurement seemed to be non-negotiable norms for most hub implementers. Similarly, dictates of hub strategy, operations, and goals often seemed to arise more directly from hub funders, long-held principles (sometimes ideology), and path dependencies, rather than from frequent, honest, comprehensive, and rigorous (not necessarily quantitative) assessments of entrepreneurs’ needs.
This stands in subtle but important conflicts with my findings that tech entrepreneurs in Zimbabwe (as elsewhere) derive a lot of value from hubs in indirect, ad hoc, and even haphazard ways—that is, in ways that are difficult or impossible to predict, measure, and attribute to any given supporter. I fear that efforts to coerce fuzzy and complex networking processes into quantitative metrics and targets, based on a limited and linear notion of “impact,” will mute at least some of the “mess” that is unavoidable in, and even conducive to, the generation of tech innovations.
When I highlighted these risks to a group of interviewees at a workshop towards the end of my stay, I mainly heard agreement that hub implementers need to keep an open mind, while making the entrepreneur the center of their attention. This led me to conclude that, by and large, the right intentions are in place among tech innovation actors in Harare, which is not a guarantee for success, but also not a bad starting point if honest conversations will continue to be had. Within all the uncertainty, it was clear that the city and its diverse range of actors and organizations would continue to be an intriguing case for the study of innovation networks.
 Thanks to David Behr for his correction and additional info (see comments)!
 Related arguments about the fallacies of quantification, detecting an ill-suited resurgence of quantitative, positivist paradigms in popular and practitioner discourse, have recently been made, for instance, in the Guardian (on data and development) and The New Inquiry (on big data epistemology).
My research is funded by the Clarendon Fund and the Skoll Centre for Social Entrepreneurship at the University of Oxford.
Note: This post was originally published on the OII's Connectivity, Inclusion, Inequality blog on . It might have been updated since then in its original location. The post gives the views of the author(s), and not necessarily the position of the Oxford Internet Institute.