I’ve spent a lot of time with past and present clients talking about RACIs. These conversations have ranged from debating whether to do them, trying to do them, or (my favourite) discussing what’s wrong with the ones they already have. I figured it was about time I put my own thoughts about this common tool down on paper. Selfishly, it may help to have a clear point of view I can share when the topic arises. And more importantly, it may help those who are grappling with this topic to approach it with caution.
In summary (TLDR): I’m not a fan of RACIs but can understand why they feel like a valid response to the chaos in many organisations today. The problem they aim to solve is the right one, but they’re the wrong tool for the job. Below, I’m going to try and convince you as to why and share some thoughts on what alternative approaches you can take.
What are RACIs anyway?
For the uninitiated, RACI is an acronym for ‘Responsible/Accountable/Consulted/Informed’. It’s a framework that’s been around since the 1950s, which is designed to show the participation by various roles in completing various tasks.
It’s common in organisations to develop a RACI to try and get clear about how work happens. This can be done in anticipation of future activity or projects (i.e. upfront, and in theory) or in response to challenges in accountability and decision-making (i.e. trying to codify or bring clarity where confusion exists).
So far, sounds reasonable…
So, what’s the problem with RACIs?
In my experience, whilst RACIs aim to serve a good intention (helping people get clear on who does what) they actually do little to achieve this in practice. Going through a process to develop a RACI with a team can provoke useful discussions on respective accountabilities, but it rarely – if ever – moves their day-to-day forward.
Here are some of the main drawbacks:
- They’re a blunt instrument. Trying to clearly define tasks and assign single point accountability for things is not an accurate reflection of how most businesses operate. To give enough granularity, a RACI would need to be so detailed or multidimensional as to be unusable (I’ve seen some running to hundreds of lines of Excel). But if you keep them high level, they either hide too much nuance or say nothing really at all.
- They’re often hypothetical. When building a RACI, you’re effectively spending time making lists of things that happen and legislating in a group environment who is involved in making them happen. But there are lots of things you won’t know until you’re faced with a given, real-world scenario, including whether this should work the same or differently as the last time you encountered something similar. There are many factors beyond notional accountability that drive the flow of work (context, outcomes, priorities, capacity, capability). So why set too much in stone?
- They’re viewed as a high effort, ‘one and done’ exercise. You’ll likely have been there – holding lots of workshops and discussions to “align on a RACI”. You burn a lot of time in theoretical conversations trying to reach the ideal artefact. One of a few things can happen to a RACI once it’s complete. Either it’s weaponised (people hold on it too rigidly when the reality has shifted around it), ignored (“I wasn’t in that workshop”) or forgotten (“I know, let’s do a RACI!”). In the two decades I’ve been doing OD, I’m yet to encounter anyone who claims to have referred to a pre-existing RACI to understand what to do next. This should tell you everything you need to know.
- They’re flawed by design. Let’s talk about the categories for a second. I’ve lost count of the number of times I’ve gone around the “what is the difference between responsible and accountable” question with clients. I also struggle to see why the A and the R should be two different people. Why have one person on the hook without also giving them day-to-day control? And of the A & the R, who really gets to decide things? This is where a lot of teams get unstuck when the real work happens. Then, what do we really mean by Consulted? Consulted how, when and about what, specifically? Finally, in the world of increasing transparency, constant information and tech-enabled communication within modern organisations, Informed is a bit redundant as a category. So, there are pitfalls with R, A, C & I once you dig into them a bit more.
- They’re anti-empowerment. If you truly want people to feel accountable for things, use a RACI at your peril. The notion of a single point of accountability in a RACI is problematic when seen through the lens of how most organisations operate. A person may care about a particular task the most, but this doesn’t mean they’re the only one carrying any weight around it. Putting people into boxes for simplicity’s sake does not do justice to how broader roles come together to get things done. Furthermore, due to the above-mentioned design flaws in RACI the “A” typically ends up being quite high up in the organisation. If you want to show how accountability trickles down to where the day-to-day work happens, you need to get super granular (see my point 1 above).
- They’re all “what” and no “how”. However well your RACI is drafted, having your name in an A or R box really doesn’t tell you a great deal. The real conversation for teams should be about how that accountability or responsibility plays out in practice. What is it you need to hold that role, and how are you expecting to work with others to understand progress or make decisions? How are you going to know when you’re on the right track, and what working rhythms do you need to keep momentum going? These are all incredibly helpful things to discuss with your peers, where simply making assumptions is leaving too much to chance. So, if you’re making the implicit explicit, better spend time codesigning ways of working than debating who falls into which hypothetical category.
I still have an urge to do a RACI. What else would you recommend?
The first thing I’d say from experience is: it’s not “now or never”. If you’re feeling a need for greater clarity on roles and accountabilities, first ask yourself why. Is decision-making painfully slow? Are people tripping over one another or running around in circles? Is there tension to solve between leaders or teams? All these issues (and more) can be solved with alternative approaches. You can seek to understand what is missing and take an experimental approach to bridging the gaps over time.
For example:
- Decision rights. If you’re in the thick of a project or a big delivery crunch, think about all the decisions that need to be made in the present. Agree the roles that people play just in those decisions. How will the decision be made (by what process and using what data)? Is there a clear owner? Who else has a voice? Deciding who decides things is a significant enabler of flow within teams, as it is the absence of this clarity that creates a lot of wheel spinning. If you like, you could utilise a decision-framework like RAPID (another acronym!). Once you’ve worked through all the present decisions, leave it there. Don’t waste energy or goodwill litigating future decisions that haven’t materialised yet. Instead, come back in a few weeks/months once things have moved on and you have a new set of things to discover.
- Stewards or “informed captains”. Rather than try to parse out types of involvement from a lot of people in lots of things, try distilling your team or organisation’s stack to a critical few. For each, agree which person is best placed to steward it. This doesn’t mean they’re the only one accountable, or indeed responsible. It wouldn’t be on a “critical few” list if that were true. What stewardship means that this person is sweating whether it happens, involving or working with others as needed and using their judgement to make it happen. For significant outcomes or decisions, Netflix assigns an “informed captain”. This person is ultimately responsible for making judgement calls on the things they own, after digesting relevant data and input from colleagues (including actively seeking out dissenting perspectives). The term Netflix use for this approach is “highly aligned and loosely coupled’, which gives teams freedom to move quickly and operate independently whilst still ensuring that someone is responsible for the outcome.
- Accountability maps. If you need to explore the interfaces in a matrix or across certain roles, I tend used this approach. What you do in an accountability map is spell out – in words – the nature of the role that different teams (or positions) play across a few different categories. For example, if you were developing an accountability map for a CEO, COO and Divisional Leader, you might have 6-8 big categories such as ‘strategy’, ‘capital and investment’, ‘business planning and performance’. Make a table, where these categories are your horizontals, and the roles (CEO/COO/DL) are your verticals. Write the nature of each roles’ accountabilities against the various categories in a few bullet points. Stress test them with different scenarios. When you have something drafted, anyone looking at it can read across to understand how the roles relate to one another and/or read down to understand more about each role. My experience has shown that this is so much more enlightening than staring at a patchwork quilt of R-A-C-I.
- Working rhythms. What I mean by this is when and how you come together to progress work and make decisions. Teams without a clear rhythm that allows them to converse or interact on an ongoing basis already have the odds stacked against them. Rather than focusing your efforts in the ‘divide and conquer’ zone of RACI, instead agree where you connect. Design the forums you need to connect in the right way and at the right time. If you have the right backbone of interactions, thoughtfully designed and well facilitated, there’s not a lot that will hold you back once you’re in the thick of things.
- Even/Overs. If your team is struggling with priorities and you feel like you’re all pulling in different directions, then developing a series of even/over statements for what you care about the most can bring some positive alignment. These are a way of stating one good thing even/over another good thing. Examples we’ve developed with clients previously include: Revenue growth even/over cost efficiency, Product market fit even/over platform stability, Market share even/over margin. This statements can provide a powerful frame for people, that they then have the freedom to operate within.
…if you’re interested to learn more, you can see how we apply them to our own culture at Kindred here.
OK I’m convinced. What next?
Well, that depends on the “why”. Hopefully the examples above offer some thought starters, and there is lots of additional inspiration available in books like Brave New Work or Humanocracy. If you’re wanting some help to explore what you need, or practical support to embed it, please do get in touch. I’d love to talk more.
