On Wednesday 29th April, two generations, with two different sets of perspectives and life experiences, joined Dixon Jones to discuss the best approaches to Analytics in 2020.
The Old Guard:
Dixon Jones, Brand Ambassador of Majestic and previously founder of one of the UK’s earliest SEO Consultancies, Receptional, back in 1999.
Jim Sterne, Founder of Marketing Analytics Summit & Digital Analytics Association, and author of Artificial Intelligence for Marketing: Practical Applications.
The New Blood:
Hello everybody and welcome to another edition of Old Guard versus New Blood with Majestic.
Firstly, a shout out to Majestic who are sponsoring the event. Without them I wouldn’t be able to put these things, and I’m really pleased that they continue to support me on doing these.
They just came out with a new Keyword tool yesterday, so have a look on the blog and have a look at their new keyword research tool that’s got some really fancy new features.
We’re in analytics mode this time and honestly I don’t think I could get 4 more able analytics people in a webinar, so I’m really pleased that we’ve got so many people on here.
I’m going to allow Jill, Sara, Jim and Annie all to introduce themselves and then we’ll start the meeting proper.
So Jill, why don’t we start with you and tell us about yourself and where you come from?
Hello, my name is Jill Quick. I am the co-founder of a company called the Coloring-In Department and we do digital marketing, consulting and training. One of my ‘love-bugs’ is measurement because “if you can’t measure it, you can’t manage it” – all of that stuff that you learnt when you were younger, so yeah I’ve got a soft spot for the old analytics.
Hi I’m Sara Clifton. I am the co-founder of Verified Data, run with Brian Clifton who you might have read some books about analytics from, and it’s a basically a software that checks data quality in governance, and I feel obviously passionate about the consent and privacy part of this.
I’m Jim Sterne, the founder of the Marketing Analytics Summit, the audience created the digital analytics Association, I’m the author of a dozen books, the latest one is Artificial Intelligence in Marketing and I help companies with data literacy and analytics literacy in what used to be workshops and is now over Zoom.
I am Annie Cushing, I blog at annielytics.com, and I’m the author of Making Data Sexy.
I tried to work out what the combined amount of experiences in this room was, and it was a number that we don’t want to really talk about!
If anybody would like to ask a question, then please ask away in the Zoom chat.
We’ve got a few questions and things to get ourselves going that some people asked when they signed up.
Firstly, all of you are really experienced in analytics and you’ve all been in a fair amount of the game to be honest, and what I’d like to know is: “What do you think is fundamentally different about analytics from about ten years ago and today. What are the differences in approach?”
So a couple things you know, it’s kind of two prongs from the same fork.
On the one hand we had a lot more personalized data… like 10 years ago we still had keyword data from Google, which you know that was pretty amazing, so we had fewer constraints on privacy and things like that.
The other challenge is that now we just have so much more data that we have to churn through and really start to marry data up from different sources to get the insights that we need.
Ten years ago we would muck around Search Console data, then called Google Webmaster Tools data, and muck around in Google Analytics and you know present them differently.
Now we’re being required to pull data from lots of different sources, marry it up and tell a much fuller story.
Well ten years ago we did not have iPad traffic, or Pinterest or Snapchat or all those things.
We also had zero formal education, the people who were doing analytics were doing it out of love only.
If they were lucky they had some statistical background, but usually it was the person who did not take a step back fast enough when they asked for volunteers so they ended up at the front of the line.
We also have such an astonishingly wide variety of tools now, that we’ve broken up into specialties.
We used to have a webmaster 20 years ago, then we had the web analytics guy in the basement, now we’ve got somebody who’s responsible for website data, somebody responsible for email somebody, who is doing data engineering, somebody who’s doing data science and then there are the analysts sitting in the middle trying to make sense of all of it and the responsibility is to be the business communicator.
Before it was just ‘crunch some numbers and give me a dashboard’ and now it’s ‘we need you to be an active, important, valuable contributor to the bottom line’.
I think of a similar point of view as Jim there. I think it’s used to be a jack-of-all-trades person that did a little bit of everything and I think now it’s actually a better position if we’re looking at it from the positive angle not just the enormous amount of data mining that we have to go through.
It’s actually easier to hire people. There is someone for things are mining they’re may be someone that is a good analyst that you could find, there might be lack of statisticians in this ‘revel-itics’ world, but were moving into a bit of a maturity stage where we can find more people that can fit all of these roles because a data miner and an analyst is obviously very different skill set.
I think for me, echoing on some of the themes that Jim had said about ‘you only did it if you like really loved it’, for me I’ve noticed that you’ve gone from maybe one person in a company, to everybody having a mission statement of ‘we are a data-driven company that want to get the insights of our customers’ or something of that nature.
I think because there’s so much information and it’s at such a high volume now that you may have lots of individual people within a business getting all the data, because people love data don’t they, they’re like give me the spreadsheets, give me the reports, but I don’t think for a lot of people they’ve still matured into the hygiene aspects of it to really understand how to use it or to know if it’s clean, so I think there’s some core fundamentals of having a good data practice to be data-driven.
I don’t think that’s moved along from 10 years ago, but I think ten years ago we were still working out what we had and what we could do with it.
Okay, the next question is: “The 2008 crisis helped digital analysts prove themselves useful by educating clients and focusing on sanity metrics such as conversion rates. What should analysts emphasize now, given the situation we’re in, to ensure all those people that say ‘just give me reach and influence’ or those kind of clients, how do you get them to be data-driven?”
First thing that just come to my mind on that is to stop looking at data as a whole and start looking at it in cohorts.
So especially given our current economic situation, instead of looking at all the data, look at who are your good customers right, now who are you losing, what can you focus on, because there’s a lot of businesses that are needing to pivot and change and you can’t just pivot a whole, you’re going to have to have some insights to help with that.
So yeah, I would look up cohorts and that kind of approach.
If you have a natural increase of reach, your conversion rate will drop.
You’ve got to pick your path here and try to segment your data and try to see what influences, who am I reaching and try to define the audiences here.
I see a lot of this on a daily basis where people try to go for reach at the same time as they want a higher conversion rate.
That’s obviously great if you can manage that but it’s normally not the way it happens, so do you want to look for a higher order value or you know who’s actually making sense here, what kind of audience is driving you forward and what are the other ones that might influence them.
The question though is what should analysts emphasize to convince those people.
So I’m a just give me reach and influence kind of guy, so just give me reach and influence.
I don’t care if they’re their influences in Beyonce songs so it doesn’t really matter, not quite, but how do you convince me to not go down that road and just say to everybody “come to my site”?
So you know speaking as the old guard: in the good old days it was “oh look we have log files” and look at all this data and you know the old joke there must be a pony in there somewhere.
So there may be value… let’s dig into the data, then let’s talk about data.
That’s the wrong conversation.
If I want reach, and I want frequency, and I want response rate, okay… but toward what end?
Let’s talk about business goals and right now you are rethinking, as Jill said, you’re rethinking what your business model looks like, what your business strategy is going to be, so what new KPI’s do you have?
Let’s not talk about reach, let’s not talk about traffic, let’s talk about how your company is going to succeed and then let me as the data person go into the dark room spin, around and come out with some answers for the business questions, rather than just give me numbers, give me numbers.
The idea that that you need to have is to understand if that is that really your goal.
If you are driving a blog, or if you’re sending assume events, then maybe that’s that is what you need to do, you need to find new customers, maybe you’re new to the market, so it’s part of a business strategy.
I’m assuming that you want to make some money on it so it’s back to the dialogue that Jim had – there’s got to be a KPI set for the actual target that you set out to do.
I think a lot of this also comes down to segmentation.
So in the early days you know we were primarily looking at data in aggregate and I think a lot of organizations still do that.
I worked with a very large company they had this data that Google had pulled together for them because they were that big, and they just like created these buckets and no one question these buckets, they really relied on these buckets of keywords – but no one questioned it.
It wasn’t until I created this interactive tree map that they realized that their car sharing, this was a tax company, and they had this car sharing services bucket and it was ginormous – much bigger than it should have been.
And they found out by drilling into that data that they were tracking, this was Google Trends data, they were tracking the keyword Uber which had nothing to do with their business. It didn’t drive sales. Someone searching for Uber isn’t looking for tax related information. But they have been making decisions on this data source that was just filled with dirty data.
I really like that Jill touched on this a little bit – data clarity and I mean that drives the ability to segment to look at your data in segments and you absolutely have to do that with when you’re looking at conversion data.
I tell clients anytime you’re putting money into anything or significant resources, you need to be able to segment down to that particular source and what that comes down to is ‘is that data clean’ when we look at our channel data in Google Analytics or in Adobe Analytics do we even know that this channel that’s for reporting revenue and things like that that, that channel is even clean so you know as so I think the devil is really in the details.
I mean like your boss in one other thing about when you’re pouring money in something else that might help – remind people of the cost of answering their question so how many people came to my website on Thursday and then looked at these three pages?
Well first of all is it okay with you if I spend ten thousand dollars getting that answer for you, and if you’ve got the answer what would you do with the answer? if I could just give you the answer for free could you take action?
So remember that collecting, cleaning, analysing, and coming up with an answer to your question has a cost to it. It’s not free. It’s not just off the top of my head. It’s work. Are you willing to pay for that work and can you put it to use?
Doesn’t that create the very problem that creates a barrier between the “I just want to get it done” kind of a typical CEO of an Internet companies start-up attitude of “I want to go, I want to go, I want to go” and then the analyst who is saying “yeah but we’ll find out later, we’ll find out later”.
I mean one of the interesting things about the pandemic that’s going on at least from the UK perspective is that the governments are putting forward the point, very effectively I think but I’m probably going to get shouted down by half the country at least, that we won’t know how we’ve done compared to whether the other country until all of the deaths have been calculated from any source after the whole thing because people are dying from all sorts of other things at the same time, different countries are blaming it on you know assigning of different deaths different things, we’re not even counting the ones that aren’t dying in hospital there’s all sorts different things and it’s only going to be years from now that that information becomes clear as to how we did compared to all the other countries but isn’t that too late for most businesses?
You know if you say how much does it cost to do that and I think the problem is that if analysts take that view then nobody ever bothers to do the analysis.
Of course there’s this natural balance in between.
If the speedometer on my car tells me how fast I’m going only after I have arrived, then it’s not useful.
So yes I need some real-time reporting but there is a cost. So the ad-hoc questions the “Oh couldn’t you just add another column to that?” yeah of course I can, but here are the things that I won’t be able to answer because I’ll be busy doing this, and here’s the cost of it, and if that’s ok with you I will do the work, I will hire some outside analysts to come in and help me if that’s ok.
I’m not saying no I’m just saying please realize it’s not free.
I think for the conversations I’ve been having with some of our clients have been focused on trying to please the people at the top.
Like there’s one company, it’s a global company that have a crap ton of money, and the main KPI that they’ve got on the dashboard is bounce rate and I’m like “oh God no!”
But sometimes you can’t move things, so it’s a case of having some flexibility so if you have got people shouting I just want reach and volume metrics.
Trying to have a conversation of what are you really trying to get here like “do you want data for the point of data” or “are you trying to ask a question” because sometimes I think they don’t know what they’re asking for and they’ve fallen into “well this is what we always had, we were told 10 years ago it’s bounce rate why should we change” or you know something like that.
I think having a look at what’s the real question they’re trying to answer, because sometimes it’s not the thing that they’re briefing you to do it’s a different problem that you need to address.
And some sort of flexibility, we’ve tried pairing metrics. So if you have got somebody insisting on “squiffy” metrics, trying to pair them with something else along the journey to kind of show the impact of it has been helpful to find a compromise for people that are still adamant about not changing some bloody dashboard because they like it the way it is and make big decisions on it.
So the client chooses a metric and then you choose a pair metric?
But one that does genuinely impact it, one where you can see some causality hopefully.
I mean all of that leads on to the next question “how do you instil data discipline, i.e. when you join a team collecting lots of data but it’s not being used correctly or not even being collected correctly and that must happen?”
That’s an excellent question and I’m going to go back to ensuring that the data is clean and it’s easy to talk about that. One thing I’ve done across projects – I’ve instilled in every workflow before I touch the data the data goes through a data prep tool.
I you worked with tableau prep and Google’s data prep with bigquery, but at this point now I just won’t work with data without having some kind of tool where I can get that 100-foot view and know that the data I’m reporting on is what it says that it is.
Because in these tools you can see an overview of oh well you know what 17 thousand rows aren’t showing up because for these 17 thousand rows a date was formatted as “month month day day year year year year” and the rest of the data source was “year year year year month month day day”.
Or I’m finding all of these countries that are recorded differently across different data sources like one will call Russia “Russian Federation” and another one is “Russia”.
If you don’t have some kind of data prep tool that you’re working with to see at that granular level you could be reporting on data and no one would have any idea because we’re always rolling it up.
Even if we segment we’re rolling it up by country, or by device, or whatever it is that we’re rolling it up by.
I just had this conversation this week where I was just like you know what and we’re working with bigquery, all the data is coming from bigquery, and the person who put it in there feels confident that it’s clean, but when I pulled it into you know Google’s data prep tool all we found so many issues, so many nulls, having to decide all right how do we handle nulls, do we want to replace it with a zero or do we want to keep it as a null.
So I think unless it’s incorporated into your workflow there isn’t going to be discipline because it’s going to depend on the whims of whoever is looking at the data at that particular time.
I’ve got something to add on that. I had a client about two years ago that was a great idea and I’ve tried to install it into others since.
We found that just like I have two small humans and to try and get them to do anything you have to reward them with something that they want or something that they care about.
So with things like data discipline of say tagging a UTM parameter correctly or checking that they’ve done things correctly we’ve been baking these requirements into their core competencies of staff as part of their job, so you want the promotion, you want the raise, do your core competencies otherwise you will be downgraded you won’t get the promotion.
For the agencies we started to bake it into their service level agreements. So if somebody was doing our Facebook marketing for us we’d say listen this is how we’re going to tag it “Facebook”. Not “Facey-B”, not “FB”, not whatever you fancy, and if you don’t do it correctly as per our channel planning I’m not paying you – you’re going to void your terms and conditions.
That really gets people to sit up, because it’s not a case of oh yeah I know I need to care about the data but I’m not bothered. You need a stick of some sort to motivate people and money, promotion, the things that we like out of our job is always something to us.
But they’re not always able to – it depends on the on that on the content that’s being delivered.
I put an advert on Facebook today and it was a video, so basically I wanted people to watch the video so I can’t really, well maybe I can and I just don’t know how to track that, it doesn’t end up with a click, it ends up with the person in Facebook becoming aware of the concept and the product which I figure is exactly what I wanted them to do – but I haven’t got that click into the site so I haven’t got any analytics except
for whatever Facebook gives me so you can’t always do what you’ve baked in.
Or am I missing things? Have I got to build things that are trackable?
I think you need to build measurement documents and processes and bake them as part of core competencies. So whatever that is and whatever data you’re tracking – just try your best to educate people on it.
When I started doing I remember reading Brian’s books, I’ve read Jim’s books, I’ve done all of annielytics, I love being on here I feel like I’m with the Stars [LAUGHTER] and I remember going into jobs and being like “so where’s your measurement plan” and they’re like “what?” and I’m like “measurement plan, any documentation” and they’re like “maybe there’s a ticket somewhere or an email?” we’re not there yet.
But people care when they lose money so in times of crisis, now people want to know what return on investment they get, so now people care a bit more.
I think you’re human and that’s why you make these mistakes.
Whether you’ve been in an education and workshop a hundred times about the UTM campaign tracking tools out there that helps you with this, you’re still going to do your day job a little bit too fast and you’re going to miss out on these and then analytics spreads across different agencies and different people in the organization.
So you’ve got to assign, in larger organizations anyway, data governance staff that is keeping on top of this and at the moment they’re the DPO’s of the world who are not heavily involved in analytics. They kind of understand GDPR and law and trying to get the grasp of what a cookie is, but we need to get them into the analytics field and understand how they control this maybe it’s a business decision to drive that as governance has to be a part of it.
I just had a webinar and there was a back-drop which had “clear on vision, flexible on detail” and I think that’s kind of me really. Flexible on detail doesn’t really work in the analytics world you know a little bit of care on the detail goes an awful long way!
If I can piggyback on something that Sara said, and I’m really glad she addressed it, is data governance.
You know because our industry is growing up and we’re having to deal with a lot more data governance and working with data governance departments when we’re working with clients or when we’re embedded in an organization, and one of the things that I recommend is when you detect an issue, and I know Jill touched on this a little bit too like with campaign tagging and because that’s a really, really big one and people don’t monitor their other channel which is where all the good data goes to die.
So one thing that I’ll do is I will just assume high recidivism rate. So I’ll weigh in on okay you were tagging the medium that you know pink gorilla and Google doesn’t know what pink gorilla is. So let’s expand your email channel definition to include pink gorilla, all even though I’m telling you to change pink gorilla to email, I’m not trusting that you’re going to do that so we’re going to expand your channel but then I’ll create just like a tech dashboard to monitor those issues, to monitor the other channel, to monitor your eyes in the content reports with lots of query parameters for ecommerce sites and things like that, because invariably these issues that you’ve detected one time, they’re going to come back up.
So when you can monitor this simple dashboard, you can build it in data studio you don’t need tableau, you can build this in Excel. I mean it’s such a simple thing to build using you know like the Google Analytics API for example, and it looks like you’re hawking over the data in the name of data governance but really it’s just a matter of checking in on this dashboard to make sure that they haven’t slipped back into their old ways.
Okay I’m going to go and pull in a question or two from the chat.
What are the biggest data privacy challenges that you’re seeing in 2020 and beyond? This was also something that you mentioned personalized data and stuff at the start Annie and then the fact we got constraints now that we didn’t have before, so I think how do you get around that?
I mean on one-level GDPR says, in the UK anyway and Europe, that you’re not even allowed to track an IP address, but you can’t deliver content without an IP address so “how does this come together and what are the big challenges, what are the ones we have to fix now, and you know who’s going to sue who first?”
Well I think first of all people have got to obey with a cookie banner consent, I think it’s 99% of all the cookie banners we see out there are actually violating what they say. They’re actually tracking you without consent anyway and I think that’s a lack of knowledge, a lack of understanding how to use web analytics.
it’s anonymous and should continue to be so, so whether you kind of keep that for retargeting or not is what you need to say to people and then people are still trying to figure out where does this fit in within my organization like we talked about before. Who owns this question and who’s actually responsible for checking it and we see anyway with clients that I work with that sometimes they’re so keen on doing retargeting anyway and staying on top of that they don’t they keep doing it until someone kind of comes and knocks on the door and says you need to change your strategy.
So they’re going to continue until it’s a very hard sharp line and lots of law cases are coming out then we might see a change, a shift.
The other part is the privacy part of tracking people when you don’t even know you’re doing it. You’re putting up a contest online and you might sort of get your data, email address, passwords and all sorts of things ending up in analytics without you even knowing it. The data that we have is similar to what Jill shared a little bit earlier when we talked here and it’s twenty / twenty-five percent of all web sites try to collect this without even understanding and they’re violating terms of services of Google Analytics or something like this, but also the GDPR.
There’s also the other challenge though that as a user the number of times I go to a website and I get the big pop up, and sometimes you just have to press the buttons or sometimes you go through the motions and stop at all, but invariably it comes to something like a Quantcast menu, so it’s the third-party tracking for sure.
If I just don’t want Quantcast the problem is the second I go to the New York Times and then click the yes button I’ve opted back into Quantcast again.
So the GDPR legislation was fairly blatant about how it’s got to be consensual opt-in. We’ve got actually to want to opt-in to this thing. But if I go to another website and click back in, then I’ve opted it back into Quantcast because I trust whichever other website run it.
Is it the data analytics collection systems? Are they the ones that breaking the law or is that each individual one or is everybody? Or am I just being too paranoid?
I think it’s a brand issue there, but I think it’s more important to talk about who wants to own this data. Is it a unique selling point to keep this data?
What you’re saying there is what maybe made you annoyed as a customer or a potential customer and then there’s got a brand reputation problem here, which is far bigger in the context.
I think that’s what’s happening to the world now, we should go to a reset mode now when we start thinking about what data should I collect and what do I have the right to do, and how does that affect my brand.
This is a cost that’s really difficult to manage because I want to collect as much as I can so I can do the work that I the best work I can do, and I’m going to use a variety of tools.
Well just auditing all of those tools for how compliant they are, is a very large task that’s an issue, a problem.
I said those are mutually exclusive. I can’t be concise and complete at the same time as that’s never going to be clear.
As a customer I want you to tell me why you are collecting what you’re collecting and it is an exchange of value.
If Amazon wants my email address and wants to know who my favourite authors are, and in exchange they’ll send me discounts on books that haven’t been published yet – that’s a service. Sign me up.
But if you want all of my history and my employment records and my household income so you can send me a newsletter about your banking options? No we don’t have a deal.
There’s has to be this value exchange – that’s the customer side. The regulation side, the regulatory legal side of it? Ooh big issue and it’s a risk management question. How much do you want to spend to protect the company against lawsuit?
So the approach I try and go with is, and tell me if I’m being mad here, and I suppose I maybe should talk to lawyers on are here and not analysts because you’re going to be biased towards ‘well we want enough information to be able to do the job’ so it may be an unfair question to put to the analysts, maybe I need a law group coming in for my next panel.
My approach has been I’m going to be collect very little data, I’m going to try and anonymize the last three digits of the IP address and not collecting the other information until somebody does something that requires them to interact beyond just looking my webpage.
So I try and avoid having the ICO pop-up, and I’m not doing a lot of those tracking things until somebody signs up and then I track them from that point. When they actively choose to sign up and tick the box that says you know these are the things I agree to.
Is that a good approach? is that relatively safe compared to what seems to have been done five years ago before GDPR came along?
I’m not a lawyer, I don’t play a lawyer on TV but if you can exhibit intent then you can defend yourself in court. We have had a series of meetings we’ve had these discussions; we have these policies in place. This is the process we’re using, if it’s incorrect we stand ready to fix it. But we are actively trying to protect people and protect data and we’re actively trying to comply, but this other tool that we are trying out on a temporary basis is collecting stuff we were unaware of, we’re so sorry, we’ll stop using it, we didn’t know is a reasonable defense.
Thank you. I think that’s a good a good place to move on from the GDPR stuff.
So a quick-fire question round for people “what’s your best book, apart from your own, what’s your best book on digital analytics that you recommend?”
I still vote for Avinash Kaushik. It’s not terribly current, but his writing is brilliant and his blog is terrific.
Sara might not say it, but it’s a good one I really liked, the smaller orange analytics books that Brian (Clifton) did.
Yes, that’s the one. I remember when I first read that it was gave me some strong arguments to have internally about cleaning data and how to get the templates and the measurement plans and things like that.
I think those lessons are still really key in terms of getting that data hygiene.
And because you need it to look pretty, obviously Annie who’s got her book on making things look less like garbage more like something I can read that’s usable has also been good because I hate Excel. So anything that can help me make things look tidy it’s been good.
Well I’ll go for Annie’s spreadsheet then. She’s got a fantastic spreadsheet for SEO in particular that I think lots of people still use. It’s almost an audit checklist which isn’t a book, but you know it’s as good as because it’s very useful and practical.
This isn’t specific to web analytics, but I’ve been consuming, I mean really kind of obsessed, a lot more with statistics because I think, and I know someone touched on this a little bit earlier, I think that’s been a piece of the puzzle that’s kind of been missing in the web and marketing analytics.
Especially with things like predictive analytics and things like that, so that’s where I’ve been spending a lot of my time.
Okay well we’re already down to six minutes left, that’s gone fast, so I’ve got a few questions to choose from here. There was a big data question, which probably with my Majestic hat on I should ask: “Of late everybody’s talking about big data analytics but a) what fraction of those people are really aware of what it is and how could it be utilized and how to convince them on the outcomes, and b) also how useful could it be in b2b services business?”
I will quote Stefan Hamal who said years and years ago that Big Data has the definition of that which does not fit into an Excel spreadsheet.
Now have the tools I mean ‘bigquery’ you can just use throw everything in there and that was the value and the fallacy. Let’s just collect everything, throw it all in a big pile and then figure out what questions we want to ask – that’s forgetting though there is a cost.
So big data has value because we can find correlations we couldn’t have seen before and we’ve got machine learning now to help us find those correlations and anomalies and outliers and that’s fascinating if you can afford it.
I think most people are not there to mine it, that’s the problem.
We’re baby steps into this and if you still don’t have such statisticians in your team then you really shouldn’t invest in the tools either, then keep yourself in spreadsheets.
I remember the early days of when I was working at Majestic, back when I was a CMO there, and a lot of people were saying “oh can I just download all the data please”.
It’s warehouses full of data! It’s physically massive, you know, and then one day somebody put up a post saying I would like somebody to build a tool like Majestic, I’ve got a budget of $500 or some of that which I thought was really quite entertaining – they took it down the end what I took out the mick out of them for it.
So last question then just before we head off “What’s your view on the proliferation of the tools and technology landscape whereas on the contrary the consumers of data or insights are expecting more consolidation. They’re expecting a more unified view of information, but it would appear that there’s much more proliferation of tools and actually the information is getting more diverse. What’s your view and how do you approach that problem?”
I think going back to some themes that we’ve talked about: What is your business model? What are your objectives? What is it that you’re trying to track?
Then with your resources and budget, what is the correct measurement strategy for you?
This may be something as simple as what used to be Piwik, if you are worried about Google Analytics or a free version of Google Analytics.
I think a lot of people jump to the shiny stuff which, they’re talking about Artificial Intelligence when really it’s machine learning. and the machines learning a program of data which is all wrong, and you go back into those fundamentals.
So as much as big data is wonderful and there’s not enough people to mine it, I think it’s seen as like a little ornament, you know? something pretty on your shelf. We’ve got the data, we’ve got the warehouse, do you do anything with it? If you’re not doing anything with the data to change your strategies, then what is the point?
It’s just getting back to basics – instil good data hygiene, data governance, making sure people want to learn about it, offer the learning, don’t kick anybody when they get it wrong because that’s how we learn. You make mistakes, you break it, you fix it, you kind of get on with it.
And although this might not be the politest way of saying it I’ve got a motto in my head which is “don’t be nasty” so if you’re thinking about collecting data, storing data, think “am I being am I making good choices here or bad choices” and that’s kind of where I lean on that.
I think a lot of it comes down to is what are the answers you’re trying to solve? What are the questions you’re trying to get answers for? and then finding out what’s the tool that will get us the answer to that specific question.
So I have a client where we’re building out dashboards for them marrying up Google Analytics and Search Console data, but then I noticed the conversation started spinning quite a bit around SERP features, so I was like well let’s pull that in from Moz because we already have these campaigns set up, so then I connected that data in bigquery and added that to the dashboard.
It’s like pre-empting what their next question is, or taking questions that they’ve already asked in a meeting and thinking ‘okay let’s add that into the dashboard’, so that when they’re picking through these different charts they can get the answer to that particular question.
I think as much as we can to get data out of silos and get it all working together, typically I found that that’s in dashboards. I think people will get much more value from the data that they have access to that a lot of times no one’s using.
Okay I’ll tell you what guys, I’m just going to do a final plug for Majestic and then let you guys just give a plug for yourselves and where people can find you.
If you haven’t used Majestic out there, it’s crawling 7 billion URLs a day and they have just come out with some really cool new Keyword Suggestion tools that’s got volumes and various other bits in there as well.
They certainly have the best metrics when it comes to Link Context as well, so do give them a try if you haven’t! So thanks to Majestic for sponsoring the event.
What’s the best way for them to find you and just give it a little hint as to where people should go to find out what about you.
You can find me at the thecoloringindepartment.com.
You can find a very good data quality tools that I’ve built with Brian over the years which checks governance and analytics accuracy for Google Analytics it’s verified-data.com.
If you want to get the book the Jill plugged so nicely it’s on BrianClifton.com or you just connect with me on LinkedIn and we’ll share some discussion there.
@JimSterne on Twitter or my professional life can be found at targeting.com
Annielytics.com – it started as a joke on Twitter, then became a brand and the book is Making Data Sexy.
That was fantastic guys, I really want to thank you for taking the time today.
Honestly everybody that’s watching and all of us we really don’t take it for granted, and we’re really grateful for you stepping in and joining us today.
Follow our Twitter account @Majestic to hear about more upcoming webinars!
Or if you want to catch up on our previous webinars, you can find them over on YouTube: “Old Guard vs New Blood” webinars.
- Old Guard vs New Blood: Rookie Mistakes - October 26, 2020
- Old Guard vs New Blood: Crawling & Indexing Edition - October 8, 2020
- Roadmap: Backlink Fidelity vs Backlink Volatility. - September 29, 2020