Skip to main content

This post is part of a learning series exploring good practice, examples, or new applications of MSD.

Thanks to Dan Nippard for this contribution.

Following my contribution on ‘K is for knowledge-sharing’ it would be hypocritical not to keep with the sharing theme. I’d like to share a few mistakes I made as a project manager and how I tried to rectify them, perhaps a later than would have been ideal, but my hope is that one person’s hindsight will be the next person’s foresight!

I arrived at MarketMakers in Bosnia & Herzegovina in early 2017 for the project’s second four-year phase. If I could have my time again, I would make many decisions differently. Managing an MSD project is hard work. Inheriting a project that has had a life before you is also tough. Project managers inevitably have to learn on the job, and funders and implementing organisations need to allow their project managers some time and space to grow into their specific project contexts. The lessons below have been chosen from a longer list to avoid repeating similar experiences and insights conveyed elsewhere in the ABCs series.

Portfolio composition – experiments and a lesson in stretching yourself too thinly

At its broadest, MarketMakers Phase 2 had a total of 21 interventions, in 4 economic sectors, aimed at encouraging entrepreneurship, self-employment and business start-up, and mitigating the disadvantaged terms of labour market participation for women and people excluded on the basis of their identity. For most of this time we had a team of four intervention managers, two intervention assistants, one impact analyst, and myself; we later added three more intervention assistants. Hindsight tells me that this was mad. The portfolio strategy was to encourage experimentation and then trim or cull entire interventions (even sectors) when the prospect of achieving meaningful change was not working out. This was primarily to manage risk. We understood that some interventions would fail, inevitably. We even ‘failed attempts’ into our logframe to help with expectation management.

Whilst a sensible strategy on paper, in practice – given our limited human resources – we were not firm or fast enough to trim or cull when opportunities failed to materialise. Project inheritance, the expectations of stakeholders, and new opportunities that came our way had drawn us in many different directions. We had inadvertently created a broad, unwieldy portfolio. We were spread too thinly, which counteracted our intent to manage risk. If you can’t give each intervention the resources it needs, you increase the risk of failure, not reduce it.

Where projects specialise – for example, in fewer components, sectors, and so on – you preserve bandwidth, allowing the most important people in your team (the intervention managers) to (i) invest in understanding their systems and build their networks, (ii) deal with the many challenges that emerge during interventions, and (iii) go deeper with their system-level problem-solving sooner. A more specialised project is also easier to communicate to stakeholders and to build a brand for. With a clearer identify, more of the right people and prospective partners come to you because they know what you do, what you stand for, and what you’re good at. Pipelines swell, clarity emerges, focus is found, and the team’s job becomes easier.

I should’ve been smarter and speedier in keeping a tighter grip on the portfolio monster. The original plan was for each intervention manager to start three interventions and to continue working on the most promising two (and/or the two that best reinforced one another). Latterly, they would each have an intervention assistant by their side. But we overshot this. The team struggled to cope with an increasingly broad scope of work. Too often I was looking for a reason to accommodate new opportunities, rather than keeping the lines clear. My advice for projects of a similar stature is to go for fewer components, fewer sectors, and be more discerning about how much you start and more cut-throat about what you stop. It is more important to leave space for a smaller number of interventions to broaden and deepen naturally and manage risk by experimenting within the intervention – not by having more interventions. For small- to mid-sized MSD projects, controlling the portfolio monster also allows you to allocate more than one person per intervention, boosting brainpower and increasing team interaction and job satisfaction. Towards the end of the phase, we narrowed down to 14 active interventions and Phase 3 opened with a more coherent and manageable eight interventions.

The long wait – a lesson in giving people a little of what they want

Despite our targets being end-of-phase targets, the “how many jobs has MarketMakers created?” question was a focus of many meetings from month 6 onwards. I tried to manage expectations: “Our interventions are relatively long-term by design… we have had to pivot a number of Phase 1 interventions that were not working out… many interventions are new… a number of them will fail… impact-level results are only expected after 18 months…” We made efforts to explain how our measurement system worked, to build transparency and demonstrate honesty in how results were reported, and to improve our results forecasting. However, half-way through the phase, we had achieved just over one-tenth of the phase impact-level target. Despite reasonable progress among our outcome- and output-level targets, our (commendably patient) funder was clearly concerned. My continual refrain that “the results will come before the end of the phase… and after” was not cutting it.

I’d been a little in denial that the funder would keep giving us the benefit of the doubt. So, we had to make changes, including some culling, so that we could focus on the most promising prospects. But one change stands out in my mind: we rethought ‘quick wins’. I had been busy brainwashing my colleagues that we should be wary of chasing quick wins, fearing that these were shortcuts or buying impact, which would prove unsustainable and not build durable systems. I had convinced myself that quick wins can erode the incentives of actors in the system to copy, react, or respond to what the project was trying to change; that quick wins only produced isolated islands of impact. But now was the time to revisit that belief. We re-examined our portfolio and tried to think more creatively about where quick wins might be achievable. The only caveat was that whatever we chose to do should not undermine what the intervention was aiming to achieve from a system-level change point of view. Our ‘no harm quick wins’ longlist was shorter than I would’ve hoped but we were able to press forward with a small selection of them.

In retrospect, however, I should’ve started with this thinking 18-24 months earlier. I had taken our stakeholders’ patience for granted. Project managers should seek to partially smooth out the classic MSD pathway to impact – the ‘hockey stick’ or J curve – to earn themselves the patience of the funder. It helps motivate the team, who are otherwise waiting too long to see their work translate into impact and can build credibility with actors in the system. ‘No harm quick wins’ can be achieved if you carefully consider each intervention’s ‘entry point’ prior to finalising its initial design. If most of your current interventions have ‘deep’ entry points – where there are many layers of step changes before your investments affect system-level behaviours and generate outcomes and impacts – then try opting for some ‘shallower’ entry point in the next wave of interventions. Alternatively, you might find opportunities in some interventions to simultaneously begin with two entry points – one deeper, one shallower. Also, make sure to estimate the time between step changes in results chains. My lesson is to not build a portfolio of ‘slow-burners’, even if you have been given the time for it. Don’t make MSD too much of a bitter pill for your stakeholders to swallow. (Post-script: the ‘hockey stick’ was realised, though later than was desirable.)

Influence – a lesson that we operate in a political playground

One great piece of feedback we received was about the difference between technical or business solutions and technical or business solutions that appropriately considered politics. The advice-giver agreed that many of our project’s interventions were well-designed and offered sensible solutions that on paper could lead to ‘widespread’ job creation but was sceptical that we would enjoy the ‘widespread’ part if we were not more politically savvy. Would a positive system-level response be stifled by political economy realities? The feedback was not suggesting that we should partisan, but that we should better understand political landscapes: how politicians think, what they want and need, and what motivates support or resistance for interventions in different quarters. Were our interventions designed and executed in a way that could garner buy-in from decision makers? Could we sell our ideas to audiences with different perspectives? This was about our interventions – and our tactics – being more credible and relevant (read: interesting) ‘politically-speaking’.

We also had become more influential and plan how to get that influence. As Julian Hamilton-Peach referred to in his recent blog (‘M’ is for…), we had to seize the initiative and protect our positions. We had more actively shape our own fate. We did this by getting out of our partners’ premises and choosing to not spend all our time in our private sector comfort zone. We better familiarised ourselves with what politicians were wanting to do and, where appropriate, built narrative bridges to bring what we were doing into their thinking. We better connected how the solutions we were working towards could help solve problems they were charged with addressing. We built government representatives and civil servants into observational, or even participating, roles earlier so our work did not come as a surprise to them. Together with domestic think tanks and research institutes, we generated useful research and analysis that was in short supply, but beyond the reach of civil servants’ budgets and civil servant. These simple but ‘off logframe’ actions, helped boost our reputation for having interesting insights and solutions to political problems. It made us worth meeting or even collaborating with and helped some interventions to get to the next level.

If it ain’t love – a lesson that some relationships are worth working on, but others should be avoided

Partners are our agents of change, our in-situ protagonists. We cannot develop systems, nor deliver our higher-level results without them. We take care how we attach and detach ourselves from partners, and to whom we should attach ourselves in the first place. However, I was not prepared for the fun and games that go on during an attachment. Partnership is far more than a pre-agreed partnership agreement that both sides chronologically work through, and it demands of your team excellent people and relationship management skills. Our offer of support to a partner is agreed at a singular point in time and comes with conditionalities to achieve inter-related objectives at that point in time. But situations evolve, and the evolution requires regular attention, reflection, and reaction.

You may love them, like them, or dislike them as people and/or organisations, but your counterparts are with you and you with them – as a means to achieve each other’s ends. In theory, partnerships are symbiotic, equal… In practice, one side often wants or needs the other more, and this is tested over time. Despite us selecting our partners, few of them are, in fact, a ‘love match’ and far fewer ‘love at first sight’! When MarketMakers entered its second phase, I estimate that about 15% of partnerships were running as planned. This left about 85% requiring greater effort and adjustment – it was common that we had to take action to massage destructive elements and personalities. Some partnerships required ‘couples therapy’. A minority had to be terminated early for reasons that could most diplomatically be described as ‘artistic differences’. These situations burned a disproportionately large amount of time and willpower for what they were worth.

At a portfolio level, partnership issues can be monitored, patterns spotted, and risks mitigated. We adapted our model partnership agreement based on lessons learned and tried to empower our partners by increasing flex, particularly around budget use and timelines. We identified ‘red’ and ‘amber’ flags that can be easily missed when getting to know and negotiating with a prospective partner. We refined our list of ‘green’ flags. The real lesson here, however, is that some partnerships just shouldn’t have been entered into. Sometimes it is better to exit or completely re-write an intervention rather than (i) cautiously accept the partnership options immediately in front of you for the sake of moving forward, or (ii) work hard to encourage alternative actors to diversify into your space. The former may be choosing the best of a bad bunch, the latter risks mission drift and ownership troubles. There are usually lots of ways to achieve impact. Be prepared to move to alternative system-level constraints if you hit a partnership wall. Perhaps it just wasn’t the right time for the idea.

Ultimately, this is a lesson with a positive ending. Worthwhile partnerships do have to be worked upon, and not jettisoned. There are both short- and long-term benefits to bothering to make such relationships work. Your partners generate the impact, full stop. Another, perhaps less obvious benefit, is the word-of-mouth reputation that you develop when local actors come to understand your project as committed, constructive, enthusiastic, and adaptive. Rigid projects do seem to acquire a negative reputation. Caring if a venture succeeds and being prepared to go the ‘extra mile’ tends to generate goodwill that a project can build on later. There are upsets along the way – not everyone enjoyed the experience of partnering with us. But, even in these cases, a reasonable proportion of the difficult partnerships endured and played their role in developing better-functioning systems and achieving an impact. No pain, no gain.

Inquisitiveness – a lesson in finding people with the humility to want to learn more

Most of us like it when the person we are talking with is brave and honest enough to straight-up admit that they don’t know something. It’s also great when that very same person follows up their admission with a quick “but let me get back to you” or “I’m on it”! With regards to project team composition and recruitment, there is a lot for Team Leaders to consider. We may not know exactly what skillsets we really need before issues/sectors are selected, constraints diagnosed, and opportunities identified. Similarly, under what circumstances should we recruit ‘jacks-of-all-trades’ versus ‘specialists’? When building diverse teams, what should we do to create the culture for diversity to thrive? For me, however, one recruitment and team building ‘rule of thumb’ stands above all others; and I believe it to be crucial when recruiting for any position on the project – does the candidate have the characteristics of a growth mindset? MSD projects demand inquisitive minds and adaptability. This isn’t about recruiting only up-and-coming youngsters; a lot of experienced people surely also possess growth mindset qualities. Though, if you do wish to recruit for experience, you might also put the candidate through their paces with regards to testing for growth mindset traits before making an offer. By the way, USAID have produced a great resource on this topic with their Guide to Hiring Adaptive Employees (if anyone knows of any other good ones, please get in touch)! Context is important; a lot of your project stakeholders may prefer to be sitting opposite ‘experience’, irrespective of whether or not it is served with growth mindset traits. ‘Experts’ often have a high currency. But projects don’t need to employ experts. We can also choose to use experienced experts as consultants. Indeed, matching an expert consultant with a core team counterpart who is hell-bent on learning, committed to problem-solving, and very helpful – often also positive, quietly confident, resilient, open-minded and ‘can do’ – can be quite a formidable combination. Having a critical mass of people with growth mindsets in the room for a team meeting is a great sight to behold. Of course, this requires that project implementers and their Team Leaders establish and maintain the right culture and take measures to temper certain ‘excesses’, channel the energy, and keep the focus. It is also important to look out for other personality traits at the recruitment stage that may be more destructive than constructive, sometimes well-hidden among other growth mindset qualities. However, I would argue that it is easier (and more fun!) to implement an MSD project with such a team. It is important that the majority of your team is not composed of people whose principal or sole source of information for making decisions is what they’ve done or seen in the past. Professional work experience may provide a head-start in many project management scenarios, but experience alone can grow to be unhelpful, restrictive, and staid if it is not deployed by a flexible person with an inquisitive mind. Team Leaders need their teams to explore, investigate, question assumptions, search for alternative courses of action, and ultimately find ways to make something (new) work.

I’m sure at least one – perhaps all – of these five lessons will resonate with many MSD project leaders and managers out there – and who could, I’m equally sure, add as many again. So, feel free to do so! If this blog is an opportunity for something of a Team Leader ‘confessional’ then it will have been worth every minute in the writing. Meanwhile, you may also be interested in an ongoing initiative under the DCED’s MSD Working Group that is looking to refine a framework of MSD Team Leader competencies, identify how new MSD Team Leaders can be nurtured, and existing ones better supported. The output will be available soon, though you can find a taster of their preliminary findings here on YouTube.