Top Data Governance Mistakes And How to Avoid Them - Nicola Askham

thank you I'm sure you're a bit like me I love conferences I love talking to people and hearing people's experiences but by this time of the day you're feeling a bit brain-dead if you are just forgive me if I am as well I got a very late flight in last night so I did I was fine because I loved the buzz of conferences but by now my buddy's trying to tell me it really didn't get a lot of sleep last night so I talk nonsense tell me don't just sit there politely so top date to governance mistakes I can't take you through all of them because I'm only allowed half an hour but if you want to know more if you go to my website there's a free report it's actually got nine mistakes so five I'm going to go through with you today and there'll be another four on there so I don't know most of you might have heard of Niels Bohr a Danish physicist and I'm probably totally misquoting him now but he said an expert in the field is a person who's made every conceivable mistake in a narrow field of expertise so by that definition I'm not yet an expert but and that probably never will be because I've made most of these mistakes but the other ones I've learnt from learning from my clients and from people like yourselves at conferences and because I now know them I will avoid them and never become an expert but I think the thing say is that there are loads of mistakes you can do whilst doing data governance and it's not gonna be easy but if you kind of keep an eye on some of these things you can't avoid them and make life a little bit easier for yourself so mistake number one right initiative is IT led so who here is running a data governance initiative out of their IT department one two three four so now people are getting a bit more brave now they've had other people put their hands up don't worry all is not lost I'm not going to point out make you stand up go look at these other people doing it wrong they've just got this slightly hard challenge than the rest of you so you can smirk a little bit so why is this a problem so let's let's ask a question question to the people who are doing for my team how's it going is it easy as its to be fair those who doing it from the business is it easy yeah so that makes you feel a little bit better so why is that why is it a problem do you think exactly spot-on thank you party so it's to do with ownership should or does IT own the data no see I'm preaching to the converted here now if I go into a company it tells me IT owns the data the first thing I say to them is say you were to design and develop a new business a new product and they go yes and I say would you pick up the phone to IT and say we're going to do a new product just bung some new fields on the system for us and they say don't be stupid Nicola but no you tell them what you want to capture about this product exactly went so you own the data it's not always as easy as that to convince them but it is to do with that so I'm not saying it can't be done from IT but it does cause you some problems and the kind of things you would get is the business is either not engaged or even worse not involved I have been involved in once where IT are just trying to do it and they're doing it for the right reasons so I don't think you're bad people I think you know the problem is i t see the problems with poor data before the business because they're normally the ones running around behind the scenes making things happen when the business have got no idea about how much effort is going in to stop the the broken data breaking systems and things like that so I T get it and never have any trouble convincing IT that data governance is a good thing but what happens if I T run the data governance program is that it can be seen as them just telling the business what to do and do the business like that not usually you know your IT you work for us you should just be doing the infrastructure I actually think that there shouldn't be of em on us that we're all on the same side we're all working for the same organization for the same goals but sadly I rarely see that normally we have a pity battle between IT and business and being the data governance person I often get caught in the middle but if you haven't got the business properly engaged you to a non-starter because they're not going to take ownership of their data they're not going to do the processes that you want them to do so it's really really hard and say IT ownership of data possess if that is the perception your organization and you're stuck with things like central cleansing so I get told oh we don't need you Nicola because we fix things in the data warehouse No so you just do default values on the way in I mean it's not quite as black and white as that yeah but you don't know that that is the right thing and we're still letting some business people do the wrong things at the front end so we're still doing this continuing cleansing a cleansing actually becomes a process they're not a one-off to fix something when we know it's a problem this is just become something we're always going to be reactive we wait until it goes wrong then we run around and find the people who can tell us what we should do to make it right and the underlying causes are never addressed as I just said so you don't go back and find out that it was a total lack of training and people are putting the wrong thing in the fields or that actually there is a problem on the system I one of my clients at the moment has got a very old system like the ones we were talking about this morning it's not quite 60 years old because I think the company is only 17 years old but they have the same main administration system that they had the day they were created 17 years ago this system corrupts some of their data so it's not anything that the business are doing wrong and it's been going on for years but there's a whole team in their IT division whose job is to keep that system running so they've got automated jobs that runs during the day they even call one of them the data fixer it runs every five minutes to cleanse the corrupted data I mean wouldn't it be easy to just fix the bugs but I'm a non techie person just saying isn't that easy so they always tell me why not so hopefully I convinced you this isn't a good thing to do but how can you avoid it well one the obvious it's business ownership of it get the business running your data governance initiative it's somebody they're responsible for it the exact level I can get some business people running it now I get that that's not always easy so if you can't do that then your other option is to run it from right but make sure you have full business engagement so I'm marylee standing here saying I'm not next but listen to me I'll tell you how to do it you can't do it from IT successfully and one of my clients at the moment I'm working for the CIO and we are doing it but it takes an awful lot of my time is spent I know 10% telling the CIO what I'm doing and 90% in the business talking to them so you can do it from my tee so all's not lost it's just even more effort you've got to really work hard to get the business engaged and working with you and to transition it to be a you one of the things I would say is you are gonna have to have somebody take ownership of the data governance framework who is in the business to make sure that it gets taken over and maintained that way so any questions about that mistake nobody's gonna make that mistake now right that's one down so mistake number two is not understanding maturity the organization and I think from the roundtables and all the questions that have been going on today it's quite clear that there are different levels of maturity here and some people are much further advanced in their data governance activities and others and I'm not just talking about data governance when I talk about maturity just data generally you know some people really are doing things manually on Excel spreadsheets other people are trying to do the really sexy exciting analytics and then there's probably everywhere else who's somewhere in between and if you don't know where you are and what's going on it's very very hard to be successful with your data governance because you could end up taking the wrong approach because if any of you were in the sessions earlier I think there isn't a standard framework there isn't a standard approach so if you don't fully understand how much who your organization is and where you are what your data challenges are your data landscape how do you know what the next logical step for you is or what the right approaches that means that communications are often incorrect because you just say things like we should do data governance because of this and people go no that's not at all why we should do it and you'd lose credibility if your communications incorrect the very obvious thing that happens is you don't actually get the reactions you want people either do nothing which is the most usual one or sometimes I've even had it where people do the wrong thing for the right reasons they believe they're helping and trying to further the cause of data governance but it's really it's just so frustrating because you've had this opportunity and you just got it slightly wrong and I will hold my hand up to saying yes I've done this a few years back you know I've merrily babbled on about what you should do and wondered why people didn't do it because I didn't fully understand what was going on at the company at that time so how are we going to avoid that one well I think it's really important to say assess the level of maturity now you can do that through you know data governance maturity capabilities there are loads of them out there and very sorry xx consulting firms here but you don't need to apply consultants this is the host of free stuff you can go on the internet if you want it depends on your organization and what you want to do but just assess where you are what are your current capabilities what and then where do you want to be because then you can plan your communications and even your approach to how you're going to do data governance accordingly I think it's one thing that people always miss are always not always but often is people don't define why they're doing data governance so in the early days of doing it I would come to conference like you're I would bounce back into the office the next day because I love learning and always ready to change the world the next day and I'd be saying I want to do this and they go why again well that person the conference said it was a good idea right and what did they know about our organisation you're going oh ok fair comment so you need to understand why your company is doing data governance and I'd love I said earlier that I'm no longer a data event data governance advantages I am a little bit deep down inside I want you all to go and manage all data properly but I've learnt to accept every use you're not going to do that it's just impossible so let's just start by doing data governance on things that we can and the things that count the most important data to your organization but you're still going to need to define why so things like GDP are actually quite useful at the moment this very some missed it respects but they're also giving us a driver that we can get people to be interested in and talk to us so that gives you a why and there's probably loads of other wise in your organization that you need to find and articulate and I think I was talking about it on one of the roundtables earlier it's you need to align while you're doing it to your strategy because if you try to either get ahead of what your strategy needs or lag too far behind you're not adding any value you need to be helping people deliver your organization's strategy now so look at your strategy and the objectives that goals that are in that and look at what your data is is it good enough to meet those things if it is great well then that's it's great in some ways it means that you're gonna find it harder to find a reason to sell data governance but where it isn't good enough that gives you a really good story to say this is what we need because we're not mature enough we can't do this without data yet oops so that was mistake to is anybody any questions on that one you're looking is tired if I feel so data governs as a project and this is a mistake who here is doing a data governance project yeah this is good somebody are slightly wavering heads hey why do you think I say this yeah I don't I don't want project to stop so projects have a natural conclusion they deliver what they're going to deliver and they move on data governance doesn't work like that as I mentioned earlier it's a cultural change as much as anything else and I don't want anybody to think that Mad niccola's stopped going on about data government so we can go back to normal it's not a case of the data governance project is finished we can you know ignore all that now this isn't just a change in how we do things so calling it a project jasc is that expectation that there is an end day and you will stop bugging them after that end date and we don't want that we want people to accept that this is just evolving to a new way of working which is why data campuses project doesn't work so what kind of things do you get the you know go wrong well you often get the checklist approach what do we need to do for data governance here's a list I'll tick them off and we must be all right so I went to a conference a few years ago in London and I met a lady who it was only at last day and she said I've been fantastic it's been amazing we've got so many problems with our data and I'm full of ideas and I'm inspired to go back and sort things out and it's just been great and I mean odd brilliant I'll say you know so chuffed that you've you've found such inspiration at the conference I bet her the following year at the same conference and she said oh it was rubbish what I learnt last year oh that's the change that's not what you said when you were leaving and she said I went to one of these sessions and it said we had to have data Rona's and I went back from the conference I went you you you you your data owners sort the data out and nothing changed no that's strange about isn't it but that's what happens when you have a checklist approach it says do this this this this and this and nobody's worrying about the why you're doing it and what outcomes you're trying to get from it so if I could write your check list and if I'm lucky 50% of it might apply to you but chances are it wouldn't do because I don't understand all your companies I don't know what challenges you've got so you need to you know get away from this checklist approach if you write you've run it as a project quite often the cultural change side of it tends to be ignored because projects are very much focused on delivering a thing usually a system or an application so if you do data governance as a project people view almost like your data Commons framework as the end result the deliverable from your data governance project and you go there you go nice pretty picture of a framework and do you think it survives if it's just being kind of we you know we put it in place and no he's going to do any support of it or anything going forwards most of this is about the users not even the the data owners and the data stewards in your data CalPERS framework these are the everybody in your organization who are producing and consuming data and if we do it as a project we tend to ignore all those people because no one project wants to talk to everybody in your organized and the one big thing I see is that it just doesn't get embedded the project's over and done with the project Mountain moves on to the next project what was working quite well at the end of the project just tends to dwindle because nobody is focusing on it and you haven't had that proper transition into BAU and you will get some some benefits it would be wrong for me to say you don't you will get some benefits you're probably fine some data quality issues and you'll fix them as you go but they're going to be short-term because over time people will go back to their old habits they're stop doing the new things that you've asked them to do and so you've had some benefits but they won't last and we all know if you do something once and it doesn't go well how do you think the second attempt is so unfortunately I've got a reputation now for going in sorting out when the first or second attempts went wrong but it means the subsequent attempts are even harder because not only am I now what to convince you that data governance is a good idea you're gonna tell me exactly why it won't work because what went wrong last time and you can be even harder to engage so it's really worth putting in the huge amount of effort to get it right the first time because it's even harder every subsequent attempt you try to do it so how we going to avoid that well I think that one's quite obvious isn't it it's the opposite of it don't do it as a project so I'm saying that but it is we do kind of need a project so kind of you heard what I just said I'm not gonna just go into a project do a project to design and implement your data governance framework you don't even have to formally call it project some of my clients have done some don't but the project methodology and approach for managing it is really really useful because it means that you have target dates you monitor and report on your progress because this is hard this takes a long time and if you're not reporting and managing it like a project then things will slip there's no way of escalating if you decided not to do what I wanted you to do this week because you other things today so you could take a project like approach but make sure you do it for the implementation only even then I try not to call it a project and I can't come up with some organizations call it a program but because I was formerly a project manager I think of a program as a collection of integrated projects so I'd think of programs not the right thing but it depends on what you'd what terminology using your organization's and so I usually call them a data governance initiative but if you've got better terms right please tell me because I get bored of using the word initiative cultural change is the big thing for avoiding that mistake is really focus your efforts on that and not on all the other bits and and overdoing the analysis side of things really make a lot of your focus on achieving cultural change getting people to start thinking about their data worrying about the quality of it or worrying about what happens to it after it goes from them and I think the one thing that people often forget and I quite often talk to people who've been given a data governance project because that's what often happens is that somebody will say oh we need to do data governance and oh you can do it and you can't no idea what data go in this is so you don't always choose the best people so that wasn't meaning you obviously you are the best person to do the jobs and one of the things I think it's perhaps key is the soft skills side of things because the act integral into achieving the cultural change and I've seen people who are really clever people they really get data they really actually understand the concept today's governance and they can't implement it for toffee because they can't talk to people they can't engage with them and get influenced and to change their behaviors so the soft skills in the person that you're relying on to implement for you are really really key well we're whizzing through I'm trying to make sure I don't run late all right not too bad so yes the last one is not going to be a surprise because I've laid with the point about strategy a little bit for a very good reason if you're not doing something that helps deliver value to achieve your strategy sooner or later something's going to question why they're paying you money to do it and they'll find something that does align to the corporate strategy for you to do instead and I know because I've been there and it wasn't data governance and I was very unhappy and that actually was the reason I left I'm a consultant so I suppose in some funny kind of way it was a lucky mistake but it was it's not nice when you've put a lot of effort into something and then somebody goes out that was flavor of the month Nikoli we don't want to do that anymore we want you to do this again but I am NOT a business architect I'm a data governance person you need to prove that what you're doing for data governance helps your organization deliver its strategy because then you can never become flavor of the month you're an integral part of delivering the corporate strategy and people will then see the value in what you're doing and it's very hard to stop something if you say oh but you do realize we won't be able to reach this corporate goal if you stop the data governance program so the implications of mistake four are you fail to address business pains if you are just doing it for the greater good if you do things because we ought to manage our data better you you might get some vibe from some people and they will tell you the things that are annoying them and I can tell you from experience that quite often the things that annoying them are probably the real niggly little things not the big things that are causing the business more pain more significant pain so if you fail to address the big pains that the senior managers and stakeholders are worrying about there they're going to perceive that you're delivering no value whatsoever at all you're just walking around evangelizing about the importance of data you've got to be delivering some real benefit for it to make a difference for them and subsequently what does that mean somebody will say well you're not actually delivering any value Nicola we better just wind up the data governance thing and we'll go and look at something else so it really is important to align stretchin that is one I mistake that I made a few times in the early days of my career and I make it now because I've got the scars to prove it you need to prove that this is helping your organization do what it as a whole is trying to do because if you don't why are you doing it you're doing it because you were like me a few years ago just you know a naive evangelist okay it's a good thing to do and that doesn't win you any senior friends in an organization really so how we gonna avoid that well you look at your corporate strategy and you identify why you need data governance so I can tell you that if you have five objectives in your corporate strategy I doubt that you're going to prove that you need data governance for all of them but you're probably going to be able to prove that at least one or two of them needs data governance that your data just currently isn't good enough to allow you to achieve those I also think it's important to identify an end goal and I was talking to somebody at some coffee break about this I think in the early days I had this temptation to tell you that you had to attain data governance nirvana you had to do everything properly everything had to be managed perfectly and everything was integrated and you know doesn't it sound lovely but do you think it's realistic so from an executive point of view I'm asking for you to give me an open check I just want to spend money and making our dates a beautiful and perfect and and that just sounds like she's off with the fairies and we really won't worry about her she's not going to add any value so I think it's important you have an end goal in mind because if you say I need to do this it might be that you need the perfect data management scenario that I've never yet seen I don't have anybody here has perhaps you can justify that for what your organization does but most companies don't need their data that good so if you can say actually we're just a bit reactive in how we do our data at the moment we wait till something goes wrong and then we run around and panic and fix it and we don't want to do that anymore you might actually be saying to exec I don't want to start a list if that's going to last 3-5 years I just want to get us to proactive so that we work out the things that are most likely to go wrong the most important data and we proactively monitor it so that we don't have to do all this reaction reactive running around when things go wrong so if you've come up with an end goal that sounds a bit more achievable you're much more likely to get your buy-in and once you've got an end goal that aligns with your strategy you can then do a road map because even with your end goal in mind it's gonna be too big believe me if you even if you say I don't want to do much when you still list what you want to do is quite scary to exert so you haven't had to think about data before so you can break that down into a road map to show them how you can achieve it in nice easy chunks and then finally today is the the failure to embed the framework and I think this was brought home to me the very first time I was a proper consultant rather than being a parent employee of the bank and I left them and I joined a wonderful consultancy that said you no longer exists and the client they put me in with fantastic great fun challenging but in a good way and was working with a good team and the data conference was going really well this sounds like I blow my own trumpet hurt them unfortunately data governance then was even more scarce in terms of experience than it is now so I think it's fair to say that I was one of the most expensive consultants on the project so the client gets the stage where they go Oh everything's going fantastically with data governance so you can roll nickel off the project cuz she's expensive now my colleagues are still there so I kind of was still getting told what was going on and I had two days to hand over what I was doing to one of their employees who had no idea what data governance wasn't didn't particularly want to do it to be fair I did feel sorry for him so he didn't really carry on all the things that I had been doing because he didn't understand them and he didn't understand the value of them and consequently the framework never got embedded and I knew about that as long as my colleagues were there but course as what happens with consultancies they moved on and I was most I suppose frustrated and disappointed a couple of years ago I was working at a client and I got a phone call from a consultancy who wanted to tell me all the wonderful stuff they were doing it this other company because it meant that they would be able to come and help this they didn't realize I was consulting this company and they went where are you and they were talking about this original client and they were back designing and implementing a data governance framework and I just had to save them but why are you doing that I did that like that should still be there and it just died the person it was dumped on with very little notice had no interest and and it just didn't get embedded so it didn't survive so the implications are not in you'll get some quick wins while you're starting to do your data governance project but you won't get anything else you won't do your cultural change because it takes time to achieve cultural change and here we go again you get a bad reputation because data governance didn't actually do anything so why would we invest in doing it again which makes it even harder to do again but ultimately I think I've spoken to people at conferences who say they do have to do it again because you might have done it the first time because something convinced you it was a good thing to do but the second time you're doing it because something like gdpr has come along and you don't have any option but you've now got a lot of resistance to get something that's hard to do the first time so how we going to avoid it do roles and responsibilities and really really focus on them this comes back to the if we just say let's all do data governance we all agree we've had a nice day today and we've all agreed days ago ants is important but we haven't agreed who's going to do it so the roles responsibilities are absolutely key you need senior people who are accountable for that data and they need to have nominated people who will run around and do the work for them whatever you call them and they need to be doing that you'd have processes for them to follow otherwise the chances of them all doing the same thing in a consistent manner are I would say slim but perhaps some being optimistic probably non-existent and it sounds really really stupid you have to implement a data governance framework on a couple of occasions I've gone in to help people do it their second or worse attempt at data governance I said so tell me about the first attempt and they get a massive you know the death by PowerPoint deck and go that's what we did before and okay no no this is just a PowerPoint that describes data governance and they're very very complicated data Commons framework but doesn't tell me what you did to implement it and I think sadly that's quite often happens is it becomes a lovely design and it doesn't get implemented so make sure you go and do something so those are my top five mistakes they're not the only ones but if you can avoid them you'll be onto a winner and you can focus on actually delivering some real value so I think it's time for the next person isn't it thank you Anna Gerlach you're a bigger part [Applause] you you

Channel: Hyperight AB

Views: 408

Rating: 5 out of 5

Keywords: Data 2020 Summit, Data 2020 Summit 2017, Hyperight, Hyperight Events, Data Management, Data Governance, Master Data Management, MDM, Enterprise Data, GDPR, Data Protection, Data Quality, Advanced Analytics

Id: WLHkeI_bWVs

Length: 28min 56sec (1736 seconds)

Published: Thu Nov 02 2017