#050: Internal Analytics Organizational Structures

Step right up! Step right up! We’ve got your org charts here! If an analyst falls in the woods, and she reports into a hub-and-spoke model, is the result best illustrated with a 3D pie chart? Join Michael and Tim as they conclude that, at the end of the day, effective communication is imperative regardless of where the analysts sit organizationally. And, because, “Why not?” ride along on a digression about the product management of analytics platforms within the organization!

Miscellany Referenced in this Episode


Episode Transcript

The following is a straight-up machine translation. It has not been human-reviewed or human-corrected. We apologize on behalf of the machines for any text that winds up being incorrect, nonsensical, or offensive. We have asked the machine to do better, but it simply responds with, “I’m sorry, Dave. I’m afraid I can’t do that.”

[00:00:24] Hi everyone. Welcome to the Digital Analytics Power Hour. This is Episode 50.

[00:00:30] So if you’ve been listening for a long time you probably have your own personal analytics skills locked down. You’re doing great and you’re welcome. Just kidding. But now it’s time to take it to the next level. We’ve got to sort out how analytics should be structured in your organization back in the day.

[00:00:53] We used to have these big conversations about where analytics should live marketing or I.T. or marketing operations and that’s cute.

[00:01:03] We’ve become oh so much more sophisticated haven’t we. So that’s what this episode is about the structure of the analytics organization. So sit back and relax as we exhaustively solve this little problem for you.

[00:01:18] Of course who better to join me on this. Orlicí into the structure of teams within the analytics space. Then my co-host and organizational expert Tim Wilson Oh I’m very disorganized.

[00:01:33] I’m not an organizational expert. This will be a fun one. This will be fun and I might go her way.

[00:01:39] So we’re at this is a good topic for us. Tim I think we’ve both of us have worked on both sides consulting organizations working within organizations. We’ve certainly got strong opinions about this. So the rest of this show will basically be US airing those opinions I think is probably the best way to describe what people are about to experience.

[00:02:00] Well he made it to episode 50 without having an opinion about something I guess get to some point. Great you are. All right. I feel like maybe one that I don’t have a strong opinion. But seriously. Why don’t you ask me some questions because I certainly do. What’s the Right. We’re going to focus primarily on in-house not agencies.

[00:02:20] Yes yes. In-house team structure because it sort of and we probably need to start creating some initial out in front boundaries or or categories of types of companies because one structure works for you know a large enterprise potentially where as it is a silly nonsense in a smaller or medium sized business and a lot of that has to do with you know well how much is going on in this company and how important analytics is to kind of the operation of the company and where it sits in the value chain and all those kinds of things. So what have you seen work. Well let’s use that as a starting point.

[00:03:00] Well so what works well is open communication regardless of what the winds are.

[00:03:04] I’ve definitely done what I can to talk about now. Sorry. Sorry you go.

[00:03:11] You know I tend to see when there is some sort of organizational structure that has the analysts is somehow relatively closely linked and not out on islands. So if you look at kind of the one extreme is every department that has a need of use of data kind of staffs finds hires promotes gets their analysts and they staff them in and they may say you know who the other analysts are in the organization. To me that that tends to be less successful because you just don’t get you don’t have the the community best practices and sharing and cross departmental benefits that can happen.

[00:03:51] So I tend to go towards a more centralized organization and that’s where I’ve feel like in my Both clients exposed to as well as being in-house when there has been the this is the analytics group or whatever it’s called the analytics department the customer insights and analytics department the analytics Center of Excellence call what you want. Having said that I think making sure this structure in the way that that gets run to have some alignment with specific business areas. And again we’re talking larger organizations here. I don’t think saying oh it’s the pool of analysts and they just kind of sit there and it’s a first in first out. Q You know when you’re done with proport X go and take the next request in the queue no matter what area it’s from. That doesn’t work either. So I don’t know that’s kind of I tend to skew towards some level of centralization of the talent but with some alignment to specific domain areas to the extent that the size and the mass makes sense I don’t know what what do you think.

[00:04:57] No I actually I’m in I think I’m in agreement for a couple reasons and I’ll add on a couple more. I think the sort of centralized structure does a couple things that I like. There are other models that I think could work. You know there’s embedded models right where there’s an embedded team inside of a group that dotted lines to a central organization. You know I’ve heard that call like a hub and spoke model or whatever but I don’t know if I’m getting that right. Hub and spoke might be something different but it’s OK we’re not making pictures. I’ve drawn it up as a little analytics fort where you build like a little structure around all the different incomings things that you have to. It’s probably not going to call it a fort because that makes it sound like you’re defending against the rest of the organization which is probably not the right thing to help build that communication thing you talk about which I agree is very important. But the other thing that a centralized structure helps with is around cultivating talent and potentially helping keep talent longer because what you know there’s inconsistencies in terms of how people work within an organization. And if you have sort of your analysts out in various different places the consistency that with which they’re all kind of learning and getting educated and becoming standardized on a common set of ways of kind of looking at and solving the business problems is less likely to happen if you find the right leader for that group.

[00:06:29] You have an opportunity to create stability across that set of analysts which then gives you a chance to keep them longer than the industry average which could pay off in big ways in terms of their value increasing over time because they’re more knowledgeable about the business as their business acumen grows. Certainly if you’re staffing with fairly junior analysts I think a really important consideration is who are they going to be working with DAY TO DAY is it going to be a bunch of people who don’t really understand what they do or is it going to be a bunch of their peers who get them and can help keep them can help them weather some of the normal business storms if you will does that mean you’re a little fort being not a good term but that it can spark in me.

[00:07:12] That’s the downside of the centralized that you can wind up with the right where the group where the group off in the corner that either people don’t want to work with or it does wind up creating separation where it’s hard to get integrated with the organization and it’s easy to actually criticize the groups that you’re supposed to be supporting by saying they don’t get it. They’re asking stupid questions.

[00:07:32] Yeah. And that that is one of the drawbacks. And that’s where you have to balance that right. I think with having a point of view of this organization exists to help support and serve a drive value to these organizations so it has to be in the charter of that group the way we’re measured is the value we create out here.

[00:07:54] And if they if they tell you that they’re not getting value then they’re right kind of definitional.

[00:07:59] Yeah absolutely or you sit down and say hey how can we help drive value for you when you know hey we’ve provided X number of analyses we’ve we’ve thrown we have position a number of recommendations that could help this team or this brand or whatever it is that you’re working with and we haven’t been able to get traction or progress together what can we do. You know when you start that again that communication piece it’s hard to get away. I know because communication basically solve like 90 percent of every problem. And then it’s sort of like well we have to do a podcast about like after that like if everyone were together most mostly things sort of fall away as issues. Luckily a lot of people are really poor communicators. So I mean I’m like wait. But that’s that going into it is why you have to think about your structure is how do you help support your organization or in. You build the most coherent team the team that’s going to be the most effective at creating additional intelligence and cross-pollinating that intelligence across the entire company. So again I think when we’re talking about this we’re both talking about larger scale enterprise you know where there’s many many different departments or brands you might be users or consumers of digital analytics. So that could be ed multi-brand companies that could be in multiday Department companies where different departments own different aspects or different applications that go to the Web. So you know lots of different ways that could work out.

[00:09:34] So we are we were kind of referring to the analytics organization as kind of this monolithic thing when it does you know we had a few episodes back where we’re talking about different roles we’re looking at job descriptions and start looking at different roles and it seems like that’s another level down even more mid-sized organizations where you have the analyst who is crunching data is making reporters doing analysis but then usually in the scope of analytics there is the technical side of implementation. There may be there. Hopefully if you’re lucky if there’s also kind of the advanced analytics moving in that data science the direction. And I don’t know that the organization necessarily needs to be I don’t know if do you feel like all of those pieces need to be in the same same group especially when it comes to the implementation aspect of tag maintenance. Taggle imitation tag management ownership of data governance you know which of those aspects is that sort of fall. Does that need to fall in the same group as the analysts or is there a case for it to be made where it’s elsewhere.

[00:10:44] Yeah no I think that’s where we get into kind of how to kind of slice and dice this because that’s the functional pieces of of the analytics or I think is where it starts to kind of break down a little bit in terms of how you structure and where you put all those things and what’s surprised me over the years is how few companies even have caught conceptual frameworks around what I call product ownership of you know things like Adobe analytics. You know it’s like well who’s the product owner Well you know we sort of own it over here in our group and it’s like well that’s probably not the best place because you can’t really do the things that other products in your organization do. Or you know well who owns tag management. Well I mean I’m an administrator over here but I have there is no clarity and that’s something that just has to be worked out. Selfishly I like you know if I was running an analytics team I’d like more control than less. But I also think it’s because I you know I’m really selfish and want to do things my way. And so I. But at the same time I have a point of view about OK what would it look like for them and what it would data governance look like. That’s not the same that’s not the same set of activities as analyzing a Web site or a mobile app. That’s a different set of activities. What does it look like to own the products that we use to do this. That’s a different set of activities.

[00:12:14] And so you know I’ve actually sort of been chatting with a number of companies over the last couple of months just in terms of our clients and things like that. We’ve gotten a chance to kind of talk about some of those things and it’s really interesting because the companies who are making progress are doing a good job of having clarity of ownership for the products that they’re leveraging for analytics so that they know okay this team is responsible for taking the tool and setting it up. You know obviously taking inputs from this group or this other people to know what to put into the tool right so that could be your implementation itself it is owned by whatever that organization is. And I think for the technical aspects of implementation it makes a lot of sense for a lot of places for that to live in an organization doesn’t necessarily make sense for it to live there every time. There’s plenty of reasons why that would be a terrible thing. In some places but sometimes it works that way. And then you’ve got you know the analysis piece. Who’s using it. How do you administer it all those kinds of things. OK well how does somebody from outside of not an analyst come ask us for help. What does that look like. Who’s taking those requests. How do we prioritize those. Who’s going to you know how do we make sure we’re kind of translating that into a valuable business question as much as possible.

[00:13:37] I agree. I mean I think that’s the hand-off from what data the analysts needs.

[00:13:42] Like somewhere between there’s this is the data that we would care about and a level of how it could be organized or structured right. The classic of I have definitely seen cases where pure I.T. peer developers have said we’ll just do the tagging what just put an insane level of detail because you can always roll it up later which in theory sounds great but that can make the data insanely messy or you could run into carnality issues.

[00:14:10] So somewhere between I don’t know if I’ve ever experienced an I.T. organization saying yeah let’s just collect all the data.

[00:14:16] No not some as collect all the data but saying we’re going to go ahead and tack on this identifier on the end of this string so that now to get the number of how many times somebody clicked on a button you actually have one row for every time somebody click because you also depended on their you know some other session identifier and they’re like hey you have all that so that I’d absolutely run into where we thought you know you wanted to capture the first three characters of this because that’s what really mattered. You could roll it up we capture the whole thing. And so you wind up with a big long list.

[00:14:50] So but from the hey let’s sort of think through how we might want to use this data which is kind of the analyst realm do we need to come up with an efficient maintainable scalable extensible implementation of how we’re doing that which falls kind of on the ITN. And I think it doesn’t really matter whether ownership of that translation lives in the analyst world or if it lives in I.T. as long as it’s owned by somebody who is able to communicate and think through the other perspective I’ve got I’ve got a couple of clients I’ve got actually I’ve got one.

[00:15:28] One client that that role kind of sits in the analyst’s world and he is phenomenal. When it comes to he’s sitting there with the analysts but he is getting stuff articulated and documented exactly what he needs to do for them in their language. I’ve got another client that’s the flip side is somebody and I.T. and he’s totally thinking through like how would this data be useful How are you going to use it let me understand that. So that what I come up with actually makes sense. But I also see places where that gap is missing. I had a client years ago where in I.T. it literally was just the ticket system saying whatever developer is available like they’re going to get assigned a task by the project manager and if it’s implement the tags they implement the tags. And the analyst organization was pretty damn weak on coming up with a good tagging schema in our approach. And oh by the way it turned out well that’s how the developers were set up. That’s about the most boring crap they could do is get shitty requirements and they’re just implementing javascript stuff and this is a pretty tag management which means they avoid it like the plague and they roll off and it’s not particularly fulfilling work and it was a mess like nothing ever went well the first time.

[00:16:43] Yeah and those kinds of structures are just set up to now work well. Right because there’s no interest or investment in the success of it.

[00:16:53] I do agree with you and I don’t know exactly what it’s called but it’s like maybe a system and you know systems analysis type of role is kind of really missing sometimes. And then is sort of what a why are we doing what we do. How does this map into our overall success or what we’re trying to accomplish.

[00:17:12] So what you’re as you’ve been talking to different organizations and thinking through so those different roles so probably the thing that to me seems like it matters the least is who actually owns the tool. The

[00:17:26] licensing re-upping contract only to the extent that whoever owns the tool also owns what’s going to come be in the tool at the end of the day.

[00:17:34] Well I guess I’m saying that aspect of it is the mechanical administrative but that next level of what the what where and how it’s being used.

[00:17:44] But even there I get what you’re trying to do but I also don’t want to use too short shrift to the product ownership side of it or product management side because done well that means the data is available the data quality has been thought about all those kinds of things are being addressed.

[00:18:00] I wasn’t I wasn’t disagreeing with anything you said. I was just saying there is a level of that which is sign the fucking contract beat up on Adobe when they come and tell you you’ve got a fucking overaged overage that they surprise you with because they are squeezing more money out of them like that. Yeah that is a pain in the ass. And is that I don’t think matters. Actual product ownership. I think you can have proud of our product management though so let’s not call it ownership. You may not like management like it and I guess that’s that was the real question is Where where are you seeing that make more sense.

[00:18:35] Well you know I think I am like you I always deviate back to my own selfish way of wanting to do that which would mean I’d want to in my short but different role. Product Management would be a different role or part of someone’s role as distinct from an analysis role or a technical role but I can see it living you know it.

[00:18:58] So in my head for making a picture for our listening audience out there Michael’s eyes are closed right now. There’s a picture of his he’s not drawn one for me so you’re in the same boat dear you’re running smack dab in the limitations of the media. Good with words. Face for five guesses.

[00:19:23] No so you know because I would see myself sitting in a marketing or research group like that’s where the analytics are would live. And in that org I could have product ownership’s there but product ownership product management sorry we now calling it product management could also I would be OK with the right partnership the product management sitting in I.T. or something like that or an ops type thing. But really we’re talking about where it is a position where does arole set that rarely exists in organizations right. I mean well I don’t feel like it’s getting less Freyr. I think it was really rare 10 years ago. You know and I think honestly I think tag management woke up a lot of companies they’re like you’re doing one other web site and that made people you know be like oh you know these marketing tags that marketing men and that’s actually what caused the problem in the first place was marketing said hey I’ve got this new technology I want to launch I t helped me launch it. Right prior to tag management. That’s kind of what was going on. And marketing doesn’t think like about product management really. Not a lot of marketers anyways. I mean if you’re a marketer and you think only about product management you know feel free to comment about that.

[00:20:45] I mean I see a lot of marketers who say they saw sales pitch and they say we should get that product and there is absolutely no ability to think how the nobody is thinking about the actual.

[00:20:56] And the plan is out and the process piece that’s the piece that you know is really needing to be managed is sort of OK well then who’s going to do what when how do we take a new idea and put it into this. And it doesn’t have to be like for every tool or one person whose job it is to manage that product. You mean a product manager could manage 25 different you know marketing and measurement technologies right. You don’t need a different guy for klick tail and GA you you only need one person to kind of make sure they’re thinking through what’s going to happen with all those and certainly tag management gives you a way to structure product management centrally so you manage the TMX which then makes you the manager of all the other technologies within it which is good.

[00:21:42] We have but that’s actually making me think through some other situations I’ve seen where it’s been I guess would be considered executive sponsor or maybe let’s just call it the person with budget authority sees a shiny bauble and purchases it and with zero interest or awareness about process or laying out a roadmap or timing or what pieces need to fit the management aspect that they throw out a product or project manager on and say get this implemented. They can just say Well my my goal here is to get this thing implemented not I’m not owning it. You know they’re a project manager and then all of a sudden they’ve got this technology out on the site and over time management help them get that implemented quickly and then they’re sent in six months in and the vendors start to ship breaks because they’re like we’re coming up to write up our contract in six months and you haven’t done anything with it because there is no ownership you might be thinking through a very specific role.

[00:22:42] It’s probably the world’s best segue into the structure of optimization teams but I don’t want to get to that yet. But no you’re absolutely right. And those in those kinds of scenarios usually have all the people involved kind of looking at each other across the table and being like yeah everybody this is you know what it’s shiny bauble syndrome and it’s on us to try to figure out the pieces you know to make you know keep this person happy who has the money that we also all live on. So yeah.

[00:23:13] Yeah. That’s a I like that. I the idea of stepping back and thinking through and I want to do that.

[00:23:21] And then on the checklist of the Midwest the best organizations have the executive who’s like OK. Let’s talk about product management and how this is going to make us better as a company. And I’ve actually been in a few conversations like that where it was kind of like you know first it was like let’s gungho personalized segment and you know Cartesian Bajun whatever. And the executive is like whoa whoa whoa whoa whoa.

[00:23:48] There’s They cartoon Asian just to see if Magar subset explodes. No I think no I’m just kind of right. I have no idea. I literally have no idea. Those are just words I have heard. I don’t know that interrupt you. I’m not going to do it.

[00:24:03] No no that’s fine because I just I said I’m to kind of indicate sort of just a you know a run by this guy run crawl walk methodology. Right. And no. But you know in that scenario the executors like well will was slow slow or roll and let’s develop something that will really be sustainable and last really sort of the organization even if it takes longer to kind of get to that point. What does that look like. And you know how enthusiastic I can be about something like that. It’s very I’m very enthusiastic because it could take a long time to get there but now you’re on a path to actually creating value with this stuff as opposed to just sort of like a couple of flashes in the pan and then everybody quits and start over again. So that’s kind of those kinds of situations are really exciting. So that’s you know let’s go back to last episode as you’re assessing jobs out there are people look for that kind of thinking at the senior levels for those out there who are realizing that Michael was actually present for the last couple of episodes he’s just a couple of episodes to go.

[00:25:11] They all sort of reach together. It’s all just one ongoing conversation. I was on them way they’ll figure out we recorded all the episodes back to back in one day and there and this is really important. You are the 3rd 2006. Very topically relevant anyway.

[00:25:28] OK so we’re we were headed and I think so we both are putting a check in the box of product management. I feel like that’s the truth.

[00:25:36] What about let me ask one quick question so do you when you’re Or do you see organizations this is back to that organization type organizations that are more pure play web like pure play e-commerce site is going to have product managers who are the product is aspects of the web or those types of organizations more geared towards saying any technology we have needs or product manager as opposed to take your hundred year old financial services company that isn’t even really didn’t have product managers internally for anything that they’re selling. Do you see that.

[00:26:11] Yeah I. No because well I mean that’s sort of where I’ve learned about product management and became a big believer in it was my good friend Kim Jensen was a product manager at Land’s End and you know he was like wait who’s going to manage this business. And we talked you know so it was it was through those kinds of conversations and experiences so I do think you have a better chance right for that to be brought up more holistically in a in an environment where product management is thought about in those terms. But I think a lot of different organizations have that you just have to tap into the right groups of the right people to be able to bring that forward call. OK. OK. Now we’re talking about the analyst organization a little bit because there’s new kinds of departments and teams that are forming out there inside of the market or a marketing realm like you know market research organization the consumer research customer insights customer experience. So where do you think where should that live inside of that. If we take the analyst organization as distinct from maybe some of the technical parts of the organization that we’ve been talking about so far in your mind Tim is there a best place for that for that analyst group.

[00:27:32] So I think I’m understanding the question. I’ve never fully understood as common as it seems to have optimization in one group that may be very separated from the web analytics group which is certainly very separated from the market research or consumer research group because relatively early in my career I worked somewhere where those were all lumped into one organization and it seemed so obvious and natural. I struggle with teams that have that split up and I’ve asked people to kind of articulate it and whatever the articulation is hasn’t ever really stuck.

[00:28:12] Maybe I just haven’t been receptive to that and it sounds like it’s back to those people you just disrespect their opinions now.

[00:28:21] Well I mean I think so. I feel like sometimes it’s like well it just kind of it actually goes back to being a product thing like the the Web and all its stuff was around longer.

[00:28:32] The consumer research market research or the competitive intelligence just grew up in a different area. You know it was the brand team as opposed to the group that was standing up the Web site. They did just kind of there were groups that have grown up in different places and I literally have not ever really figured out exactly why they don’t come together I watch now. I’ve watched cases where people removed from all of those have gotten confused and thought that somebody who was in the analytics organization was actually in the customer research organization. And I’m like those two organizations are nowhere near each other and it’s been a battle to get them to work with each other but good on you. You know somebody who’s just out there they can all ship these are the people who have data and insights I assume they’re all part of the same organization and they’re they’re not. And it just makes no sense to me whatsoever because every one of them is going to get a question that they get asked they’re going to see their little domain of tools and they’re going to try to answer that question and it comes in where somebody asks a question about what do our consumers looking for when they come to our site or what are consumers most want from our Web site.

[00:29:41] Like damn it I’m going to find that web analytics like we’ll shit. Maybe you should go put that in your next consumer survey right and ask them. I’m sorry but he has their little hammer and they see everything as a nail. Whereas if you go the one time I worked as a relet I mean this was like 20 people told all that covered the data warehouse the market research web analytics be-I. So maybe that was half a billion dollar company. It was like yeah that was our little community. People came and asked us a question and we were 10 feet away from the person who might have the best answer if it wasn’t us. And we were kind of forced to be aware of this as the broad set primary research secondary research a test a b test surveys e-mail surveys web analytics. That to me seemed like it was so functional and it is I have yet to see whether it’s rare to find other organizations that are set up that way.

[00:30:39] Yeah. And there are a bunch of different disciplines that you kind of touched on and they all kind of have their own spot. And I I struggle actually with this a little bit and I honestly don’t know that I’ve got a really clear. I mean in one sense if you’re in an organization where each of these roles is represented by like one person then by all means at least put everybody together in the same place so that that kind of conversation and communication can happen. So in other words if you get one market research person and you’ve got one UXO person and you’ve got one in a list let’s make sure they’re close to each other. Let them all talk about the problems of the business together and combine their knowledge that will benefit the organization greatly when it grows into disparate teams. You know they’re kind of running their own programs and things like that. I kind of can see them kind of going their own way but still there being some centralizing kind of structure whether that be that they all come up into one big sort of you know central structure. And I don’t know because you know in certain places you have sort of concepts of revenue management kind of driving analytics some you have experience driving analytics. So it kind of depends on the organization and kind of where analytics kind of things for that organization in terms of what they really see it doing. And yeah I guess I’m I wish I had a better answer on that. But I I can see it happening.

[00:32:08] But I do know there’s a ton of value in seeing in getting more communication happening across those disparate teams.

[00:32:16] Yeah well I think there’s also you know earlier we’re talking about when somebody newer Jr. comes in I think it’s incredibly useful for them to actually have some exposure to what those other groups are.

[00:32:28] I feel like people who are our age who maybe wound up in one area or another it just doesn’t like primary research.

[00:32:35] It just does not cross in secondary research prior research definitely crossed my mind. Hey we could we should be asking people this secondary research and what I was. One of the agencies I was at. We had contracts we were paying good money with these secondary research organizations and the people who the product owner product managers of that secondary research were like we have this valuable stuff nobody’s coming and asking us the questions we can use to answer it. And we actually pulled them in and we said hey you know what we’re going to our kind of monthly sort of get togethers where we’re sort of sharing what we’re doing we want you guys to come and share with us and we’ll share with you guys because we should all be thinking this is like this full set of tools that we have. So I guess yeah if organizationally they’re not they’re not one organization if they’re not even rolling up within one centralized organization that’s I guess shit. It’s another case for communication saying well those are. We need to know who they are and what they’re doing and who should be taken to lunch and saying what’s the what’s the core data you have and what’s that good for. Because it’s probably they probably already have it and it’s free and it’s free to piggyback on and hey you can add more value to your organization when you use the right data to answer the the question. Right.

[00:33:51] And actually that’s the kind of conversation that gets you in pretty good with those teams. It’s like what challenges have you been trying to get other people to pay attention to that they aren’t. And what can we do to help with our data set yet.

[00:34:04] And how can we route people to you if they’re asking you the same question and I copy you at all times I think you might.

[00:34:09] Let’s get together yeah yeah what’s the right way to loop you into these conversations so you can get a side table because invariably everybody’s dealing with their own set their own aspects of having buy hand or not having buy in from various parts of the organization and trying to solve that problem. So on the you know whatever our counterpart is in the market research world if there is one to market research our hour it makes you want to go look and see if there’s actually like any market risk.

[00:34:42] I’m sure there are market research pyrolysis Yeah I’m sure good on us really doing our homework on but we knew it was going to go here. No not exactly but. But

[00:34:53] to that point is like I imagine they have similar challenges. A lot of people really understand what they do or how they do it. And those kinds of things so bridging that gap and understanding the mutual challenges is something that to be really effective. OK. So we’ve touched on that a little bit but we haven’t gotten into it which is the next thing which is optimization and that sort of optimization is sort of that middle ground between analysis and execution insights right where it’s learning validation all those kinds of things. So what have you seen work there and is it the same as analytics different.

[00:35:31] I honestly haven’t seen a whole whole lot there. I don’t work with a lot of organizations that have really built out optimization heavily.

[00:35:43] It winds up the ones that the organization I’m working with it’s kind of starting in the analytics organization which I think is fine except it’s usually under under-invested in I to get back on to explain to us on the optimization front seriously.

[00:35:58] Now why didn’t we ask him more questions. No. Well it’s a good point. It is hard to find organizations that are doing a good job of this. I do think it’s distinctly different enough from what an analyst does that it may or may not need to sit into it in a different group or its own group. However I don’t I agree with you. I don’t like taking optimization or our testing. I don’t like taking those things away from the analysts. I think that’s the best and funniest part of an analyst job is that kind of work. And it really sucks when you’re like yeah yeah you got us to the 20 yard line and now we’re going to hand it to the real stars who are going to run it over the over the into the end zone. So because that’s the value creation piece of analysis analytics that’s the part where I want to keep analysts involved. However executing analysis versus executing optimization is a much different set of tasks and parameters. And so you have to incorporate or you have to account for that. And so they’re split.

[00:37:05] You’re actually are affecting the user experience. There is development changes. There is a much higher level of QAD not missing this data. I’m fucking up the customer experience right.

[00:37:16] And so there’s lots and lots of different stakeholders and groups and that’s where I’ve seen stuff like steering committees or councils kind of become really good vehicles for driving optimization forward because they help with execution. They help with follow through and they help with prioritization which are three critical things I think for any optimization program and follow through meaning accountability so think if you found something you we ran and test an A and B was the winner by you know a mile and a half. But you know now I.T. is saying well we can’t make that live on the Web site for another 18 months. You know that committee sitting there and being like the loss of opportunity for not prioritizing this in terms of execution is this much should we then use our combined strength to go and push this to the organization more effectively. That’s what I mean.

[00:38:10] But yeah I am going to back up a little bit and say that although not not clients organizations I’ve worked in.

[00:38:16] But I can think of one retailer here in Columbus where the optimization is totally rolls up it is one manager that is doing analytics and optimization but then I’m thinking of other much larger organizations where there are so many tests and they’re pretty proud of how the volume of the throughput of tests is that really the testing organization is this massive group that is largely just being fed the kind of product management product management of the site product is what’s driving the tests which makes sense so being aligned with product managers. If you’ve got strong product management of site functionality there’s some level of saying yeah they’re the ones who are accountable for making the the product of the checkout process or the product of the catalog work effectively. So it makes sense to be well well aligned with them. But yeah that whole idea of analytics and forming which test to run gets messy.

[00:39:17] Well I think it’s not just the analytics that drives that. What test run conversation is that committee because it’s all those people are like hey I need optimization over here. Can you guys because the organization is not an island unto itself right. It owns driving business value forward. So it’s going to want to look at and solve the biggest and most valuable problems if it’s doing its job. Yeah. So the large organizations that we got that one solved for fun don’t work in a larger organization. Just go work in a large organization so this will all make sense.

[00:39:53] You know what.

[00:39:57] And in reality it is because you can do so many different ways. And the nimbleness of a smaller organization just then puts everything on the people themselves. And so if you get a great person a great group of people there’s nothing you can’t accomplish in a smaller or faster growing organization which is a wonderful thing on its own. And you know there’s there are things that can scare the crap out of you like oh yeah I can change the website right here for my desktop versus you know it being kind of to locked down and how it’s got to go through these four committees and it’s taking forever for me to happen. Well our committee on when to end the podcast is telling us that now between now and the next two and half minutes is the optimal time to start wrapping up. So we’re going to do that. But if you’ve been listening we definitely want to hear from. But before we do that let’s go around and do a last call. Is this something that I show we like to do talk about something we found that’s interesting lately. What do you got for last call Tim.

[00:40:57] I am going to defer to you because I think yours is going to be like serious and useful and mine is going to be whimsical and I might have to check one a little fact.

[00:41:04] Oh I’ve got mine is actually a joke that I heard. No it’s not. A little while back I saw a tweet from Jeff Chaisson somebody who’s former d’Este on this. Yeah actually yeah. And actually in line with what he was a guest on the podcast for which is technical skills for analysts he tweeted out this link to 10 tips for maximizing the chrome developer console. And they are outstanding. These are excellent excellent tips. And where would you want to you know where do you interact with the chrome developer console. Typically when you’re doing quality assurance types of things you know testing tags trying to see what’s firing. Certainly if anyone ever uses Adobe DTM right you get debug mode turned on and you can see what DTM is doing. What tags are firing in the console is a very powerful tool and knowing how to use it is a really great thing. So not everything in there is necessarily applicable but it was a really good set of tips that I thought you know people can benefit from.

[00:42:13] That’s awesome. So since you went with such a fantastic and useful and practical I’m going to go completely off the deep end on a random and it really was inspired by our discussion before we started recording. I am going to recommend from 2005 the comedians of comedy available on Netflix. Patton Oswalt Zach Galifianakis Brian post saying yeah that pronouncer and Maria Bamford fantastic comments and they basically just do a tour and they record it and it’s kind of awesome.

[00:42:46] So that’s a little bit I was inspired because when you were talking about Maria Bamford’s new newer series.

[00:42:55] So you know go have have a laugh when your organizational frustrations have really gotten to you. Go watch the comedians of comedy tour.

[00:43:03] Isn’t that wasn’t that Ricky Gervais’s catchphrase from the show extras. You have an old laugh. Wow.

[00:43:11] Have been I saw like three episodes of extras. Actually Ricky Gervais says and I know the premise of extras and I’ve actually seen a couple episodes so about as close as you come to getting a pop culture reference ever. So I’ll take it and go with it. What was it. That’s right before the end of the year. I got it knocked out of my personal KPI for that stuff.

[00:43:35] Or 2016. Right. If you’ve been listening and you’ve got something to say we’d love to hear from you. We’re on the slack all the time. Our Facebook page you can even reach out to us on Twitter. We’d love to hear from you if you agree or disagree. And well and you know also it’s getting close to the end of the year if you’ve been a listener for a while and you enjoy even an episode or two. We’d love for you to rate us on iTunes. Let iTunes know which I think at the show but we’re does is very appreciated.

[00:44:07] It’s like your holiday gift from us. That’s right. In the holiday in lieu of whiskey we would take an iTunes rating. Now wait a minute. There you take those. Do it. Yeah.

[00:44:21] As always Tim. Great conversation. You know always thought provoking Cruciatus. And if you’re out there listening keep him alive.

[00:44:35] Thanks for listening and don’t forget to join the conversation on Facebook Twitter or MetroWest lightweight. We welcome your comments and questions. Dot com forward slash analytics analytics are.

[00:44:50] All smart guys. Want to believe as they made up a term for analytics. Linux. Don’t work. I’m ready you ready. ARE YOU TALKING TO ME. ME ME. ME ME ME ME MA MA MA MA MA MA MA MA MA MA. Would just stop there and go and talk about this next weekend. I mean that was sort of me playing for time which you then shut down. So. I’m now going to be in therapy. The fact that I just can’t get out on the. Road. We used to have. A big conversation about whether to live in marketing or to. I. Don’t know why do we. Listen. I’m not going to get that out there for you. No it’s just. So what’s your favorite. My favorite if I see code like it’s seven. Oh yeah. Validate the alternative finance insurance and real estate. So oh good lord that rain. We’re officially complete. You know we are. This is right where we want to be. I actually had somebody today Joe Flint you can and I work with them or like to do a better job talking about. You. Don’t have to go back to the. Culver City California. Just wow. Flag an org chart.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Have an Idea for an Upcoming Episode?

Recent Episodes

#243: Being Data-Driven: a Statistical Process Control Perspective with Cedric Chin

#243: Being Data-Driven: a Statistical Process Control Perspective with Cedric Chin

https://media.blubrry.com/the_digital_analytics_power/traffic.libsyn.com/analyticshour/APH_-_Episode_243_-_Being_Data-Driven__a_Statistical_Process_Control_Perspective_with_Cedric_Chin.mp3Podcast: Download | EmbedSubscribe: RSSTweetShareShareEmail0 Shares