One of the things I always loved about Fawlty Towers happened before anybody had said a word. At the beginning of the programme there was the hotel sign, initially displaying the reassuringly ordinary Fawlty Towers, but increasingly subjected to the attentions of a disgruntled paperboy as the series went on. Letters disappeared, others moved around, and before long Fawlty Towers had become Warty Towels, Watery Fowls, Flay Otters, Fatty Owls, Farty Towels and several others.
For the image accompanying this article, I've borrowed the original sign joke and continued it with an anagram of my own: "Testy Warwolf". And, before anybody reaches for the Scrabble tiles, it really is a letter-for-letter anagram of Fawlty Towers. "Warwolf" isn't something I've invented to make the letters fit either. It is recorded as an archaic English variant of werewolf (as well as being the name given to a mediaeval siege engine.) I particularly like testy. It's difficult to think of a better word for Basil, who can go from mildly irritated to apoplectic because somebody has ordered dinner. I do have to admit, though, that spending time devising anagrams of Fawlty Towers probably isn't the most productive thing I've done this week.
The Warwolf in the picture isn't there just to make the anagram work.. Basil employs people and then seems permanently astonished when they attempt to do anything without involving him. He careers from reception to dining room to kitchen, interfering with Manuel, arguing with Polly, trying to keep things from Sybil and taking problems that are already going wrong and adding the one ingredient they really didn't need: Basil. Hospitality wasn't an obvious career choice for somebody so easily irritated by guests, but management doesn't look much safer. Manuel has been at the hotel long enough to know at least some of what he's doing, Polly is generally more sensible than everyone else put together, and Sybil largely gets on with running the place. But Basil still needs to be involved. You do occasionally wonder whether Basil has ever considered attending a management course. Although I suspect he'd spend most of it explaining to the tutor why everything they were teaching him was wrong. Yes, Basil is a bit of a monster when it comes to micromanagement. The furry hands and ears are obviously a joke. Basil's management habits become less funny, though, when you start recognising versions of them in real organisations.
Some of the most extreme examples I've encountered came late in my own career. By then I'd assumed I was senior enough to be trusted to make decisions. After all, I had been made responsible for national services, large numbers of staff and analytical work that mattered. Yet at one point I was expected to copy my manager into every important decision that I made. Yes, really! These weren't exceptional decisions involving millions of pounds and the possibility of an appearance on the front page of the Daily Mail. They were the normal decisions that came with being a senior director.
I had been made responsible for national services, large numbers of staff and analytical work that mattered. Yet at one point I was expected to copy my manager into every important decision that I made.
There was a period of time, as a senior director, when I was required to report both good and bad feedback people gave about me personally. To an extent, the bad feedback made sense. If somebody had a problem with something I was doing, particularly in a senior role, I needed to hear it and sometimes (but not always) my manager did too. Strangely, the positive feedback was much more awkward. Somebody tells you you've done a good job and then you're expected to go and tell your manager that somebody has just told you you've done a good job. “Just thought you'd like to know that people seem terribly impressed with me at the moment” requires a particular sort of confidence I don't possess. There’s something peculiarly British about compliments anyway. We're perfectly happy to complain about things, especially if there’s an opportunity to do so at length, but somebody telling us we've done something well can be surprisingly difficult to deal with. Having to then report the compliment to your manager is even more awkward.
I could probably have done with a feedback flowchart.
Or maybe a spreadsheet, which brings me to those also needing approval before they could leave the organisation. Not from me, but by a more senior director. Now I've produced a lot of spreadsheets over the years, some considerably better than others, and I've seen some extraordinary ones too. Calculations nobody could explain, formulae copied across hundreds of rows without anyone checking whether the references were still correct, and beautifully formatted tables containing beautifully presented rubbish. Of course analytical work should be checked, and some of it needs serious technical challenge, replication or independent review. But what I wanted were people who understood the analysis checking the analysis. I couldn't see why being further up the management structure automatically made somebody the right person to do it. For me, it was exceptionally disempowering. After all, statistical knowledge doesn't slowly enter the body just because you're in the most important office chair.
What I wanted was people who understood the analysis checking the analysis. I couldn't see why being further up the management structure automatically made somebody the right person to do it.
So I could be responsible for a national analytical service, accountable for what it produced and answerable when anything went wrong, while still not being trusted to decide whether one of its spreadsheets could leave the organisation. Please make it make sense!
One incident bothered me much more because there were real consequences, and I can still remember the details of it clearly. I should say something about the director involved before going any further. I'm not telling this story to criticise them personally. Later on the relationship changed, and we were given far more independence and responsibility. The culture we were working in at the time is really what interests me here.
We'd been successfully delivering a national piece of work for several years when it went out to tender again. Naturally we wanted to bid. We knew the work inside out because we were already doing it, had a strong track record and, assuming we actually submitted the tender, there was every reason to expect we'd win again. People were employed to deliver the work as well, so losing it potentially meant losing jobs.
The tender was ready but required final senior sign-off, and the director who needed to approve it hadn't done so. Nobody had discovered a flaw in the proposal. We hadn't promised something impossible and the figures hadn't turned out to be nonsense. We'd been delivering the work successfully for years. We simply hadn't got the final sign-off before the deadline despite multiple reminders (and tender deadlines don't rearrange themselves around senior diaries.)
Another senior person in the organisation could see what was about to happen. Waiting meant missing the deadline and losing the opportunity to tender, so they authorised the submission on the director's behalf. I thought that was an entirely sensible course of action and I still do. Unfortunately, the director didn't. The tender was pulled at the last minute, just as it was being delivered to offices in London by courier to ensure it hit the deadline, because they hadn't personally approved it.
I could have understood pulling it if somebody had identified a serious problem with the bid or concluded that continuing the work no longer made sense. But neither of those things had happened. Somebody senior had stepped in to stop us missing a deadline and the response was to withdraw the tender anyway. So we lost the opportunity to bid, people's job roles were put at risk, and I was enormously frustrated by it. To be honest I'm still frustrated by it. It created a lot of unnecessary work and put stress on individuals that wasn't necessary. If there was another reason for pulling it, I never discovered what it was. I've certainly never managed to convince myself that watching the deadline pass would've been the better decision.
I've also worked with non-technical managers who wanted to understand every detail of a piece of analytical work before they were comfortable letting it proceed. Yes, technical staff need to be able to explain what they're doing, and I’ve very little patience with analysts disappearing into a room for three weeks, emerging clutching a spreadsheet and announcing that nobody else would understand it. The problems start when explaining the work takes far longer than doing it.
I’ve very little patience with analysts disappearing into a room for three weeks, emerging clutching a spreadsheet and announcing that nobody else would understand it. The problems start when explaining the work takes far longer than doing it.
Then there’s the manager who thinks data analysis is easy because, when you strip away all that irritating statistical stuff, it's really just about thinking logically (How many of us have had one of these?). Conveniently, they also seem to be the only person who has thought logically about the problem. Twenty minutes with the data produces the obvious answer. Meanwhile, the analysts have spent years learning their subject and may have worked with that particular data for years. They know where it came from, what its limitations are and some of the traps sitting behind the apparently obvious answer. They're then expected to explain why the answer isn't obvious at all. Occasionally, of course, the manager will be right. I've had people ask questions of me as an analyst that made me realise I'd missed something or made the problem more complicated than it needed to be. What used to drive me mad was the presumption that the technical team must have got it wrong simply because its answer wasn't the one somebody reached after looking at the problem for twenty minutes.
And I've sat through more discussions than I care to remember where something was explained, wasn't understood and had to be explained again. Then came the apparently obvious alternative, followed by another question and eventually a return to something we'd settled half an hour earlier. Meanwhile, the people who knew how to do the work weren't doing much of it because they were busy explaining it.
Of course things can be made far more complicated than they need to be. Analysts can produce a forty-page report when three paragraphs and a table would do, become terribly attached to methods nobody else wants to hear about and preserve processes that made perfect sense when Tony Blair was Prime Minister but whose original purpose has disappeared into organisational folklore. Someone without much technical knowledge can ask an extremely useful question because they aren't buried in the subject. But asking the same question for the fourth time because you don't like the answer is different.
Someone without much technical knowledge can ask an extremely useful question because they aren't buried in the subject. But asking the same question for the fourth time because you don't like the answer is different.
A lot of my career has been taken up with working across statistics, epidemiology, information standards and national health data. Some of it is difficult because different organisations, systems and definitions have an irritating habit of colliding. Years of teaching and training have taught me much the same thing. You can remove the jargon, find examples and spend an hour explaining something, only to discover from the final question that you've spent the last twenty minutes explaining the wrong thing. I've certainly explained things badly myself, but there have also been occasions when the other person simply hasn't understood something however many different ways we've tried to explain it.
I've also seen people with limited knowledge of a technical service reorganise it to make the service clearer to them. Organisation charts are wonderful things for this. You can move people around, change reporting lines and give everything a new name without having to do any of the work affected by the changes. Move the boxes around on PowerPoint and it all looks perfectly sensible. The people represented by those boxes then discover that the expertise they used to sit next to is somewhere else and the person who understands their problem is now three reporting lines away. I've then seen staff blamed for poor performance when the new process itself has made doing the job considerably harder.
Organisation charts are wonderful things for this. You can move people around, change reporting lines and give everything a new name without having to do any of the work affected by the changes. .... The people represented by those boxes then discover that the expertise they used to sit next to is somewhere else and the person who understands their problem is now three reporting lines away.
Staff would find ways around the structure simply to get the work done, and before long some of those workarounds had effectively become the real system. I've then watched another restructure come along and start the whole process again.. Sometimes I think we'd have saved considerable time by leaving people where they were and asking them what was stopping them doing the work. Admittedly that would've left at least one person with an empty PowerPoint presentation.
If somebody makes a decision and gets hauled over the coals for not asking first, everyone else sees it. Next time somebody asks first. After a while everybody asks, including people who already know perfectly well what needs doing. The same happens when every piece of work comes back covered in amendments. At some point you wonder whether there’s much point putting in much effort before sending it upwards.
I've heard managers complain that people don't show enough initiative, and sometimes they're right. Some people are very quick to pass a difficult decision up the management chain. But I've also been on the receiving end of something much more confusing. I've been told in the past that I needed to be more innovative while, at the same time, being told that important decisions needed approval and that there were certain ways things had to be done. I never did understand what sort of innovation was supposed to emerge from that arrangement.
I always tried to let people I managed make mistakes. Obviously not mistakes involving a few million patient records, but the sort of everyday mistakes where nothing terrible was going to happen and they could sort things out afterwards. People sometimes did things differently from the way I would've done them. Some didn't work. Others worked well, including some where their solution was better than mine. If I'd insisted on approving every step, I'd never have found that out.
I haven't always got it right though. I've challenged analyses, sent work back and irritated people by asking questions when they wanted me to leave them alone. And yes, there were occasions when I didn't need to be involved. It's easy to persuade yourself that you need to know when you really don't.
It's easy to persuade yourself that you need to know when you really don't.
In an earlier article, How Do We Use AI Ethically?, I explained why I've reluctantly started experimenting with question-based headings and an end-of-article checklist. I'd seen advice from LinkedIn suggesting that this kind of structure could make articles easier to find and skim and potentially improve their reach (by people and AI). Regular readers will know that this isn't how I naturally write. I tend to wander around a subject and occasionally disappear down a side road because I've spotted something interesting over there. And the last article I wrote in this way got minimal reach, which isn’t particularly reassuring. But as a statistician I know a sample of one is too small to draw any meaningful conclusions from. So I’m giving it another go.
In the interests of LinkedIn reach, here's the obligatory checklist. If you're wondering whether you micromanage, ask yourself:
Does this actually need my approval?
What would happen if I didn't see it?
Am I checking this because there’s a genuine risk or because I always check it?
Does the person doing the work know more about it than I do?
If I don't understand the explanation, have I established that they're wrong or only that I don't understand it?
Have I given somebody responsibility without enough authority to exercise it?
What happens when somebody makes a perfectly reasonable decision without asking me?
Am I holding something up?
If I'm unavailable for a fortnight, can people carry on perfectly well without me?
There. LinkedIn duly satisfied.
Thinking back to Fawlty Towers, Basil Fawlty never quite manages to leave anything alone. Every problem appears to require his personal involvement and, once he gets involved, things seldom become simpler. Having been a national Director of Analysis who needed somebody else to approve a spreadsheet, I'm no longer certain how exaggerated some of that is.
I've sat through plenty of presentations about leadership, empowerment and organisational culture over the years, and I'm sure I'll sit through several more. They tend to contain reassuring words such as trust, autonomy, innovation and empowerment, usually accompanied by photographs of people having suspiciously enjoyable meetings. I've never learned much about an organisation from those presentations. You learn it from the ordinary things nobody thinks worth putting in the presentation: what happens when somebody disagrees, who can make a decision without asking, whether expertise is actually respected, how managers react when somebody does something differently from the way they would've done it and whether people are trusted when nobody senior is watching. That's where the devil is in the detail.