There is power in data and how it can be used. Having been in the data industry and in companies such as Google, data analytics expert Gauthier Vasseur could attest to that. Gauthier has an extensive background in all the different facets of data and where they can be applied. He joins us today to talk about the importance of data management and crafting your data strategy in being successful. He also touches on the common data challenges organizations run into, what data you need to collect along the way, and what to do with dirty data.
Listen to the podcast here:
This Is What Happens Without A Data Strategy with Gauthier Vasseur
Learn Where To Begin Crafting Your Data Strategy
I have the pleasure of welcoming Gauthier Vasseur. I’ve known Gauthier for several years in a variety of contexts and I wanted to bring him on the show to talk more about data and data analytics. Trust me, it’s not going to be as dry as it sounds. You’ll understand why I brought Gauthier on the show after you hear the introduction and his background. Gauthier, why don’t you go ahead and introduce yourself to our audience?
My name is Gauthier. The reason why I was invited to this show is probably because of my background in data but also in all the different facets of data. I’ve been in operations and finance for quite a while and I saw where data could be applied. I’ve been in the data industry in companies such as Google and I could see the power of how data can be used. I also spent some time in the software industry where I could see all the tooling behind data management. Finally and last but not least, I founded several startups or been in startups and I’ve seen how important, but sometimes hard data was, but it’s important to be successful. Last but not least, I’m now an instructor. I want to share my passion for data with as many as possible because analytics need to be inclusive to be successful. Hopefully, I’ll give you the tastes of data that will make you want more.
As we start thinking about the data journey, let’s start with some of the bigger companies and then work our way down to startups. Can you talk about some of the common data challenges that you run into? What are some of the big picture, common issues that you see for organizations that are starting to try and get their arms around data?
Are you talking about smaller organizations or organization in general?
Yes, go ahead and work your way up from small to large if you want to go in that direction?
Data should never be a problem to master or to work with because that’s the domain that is not rocket science. I’ve been teaching it for a decade and people can learn and people can be proficient. The main issue for startups and for large enterprises is not going to be at this level. The main issue is first, the lack of education. People don’t know how data works. Believe me, it’s not because you’re good at Excel that you know how to work with data. It’s a totally different ball game and that’s the beginning. You don’t know what you don’t know. The other part is that many people also don’t dare thinking about how data could be useful to them. It’s much easier to remain comfortably numb in what you’re doing and what you’ve always done. As much as comfortable to say, “I’m going to challenge the status quo and get more data to challenge it.” That’s another issue I see watching small and large companies. Last but not least, while it’s most often due to the lack of knowledge and experience on data, data might be the next oil but unfortunately, it is like crude oil. It is dirty. It stinks and it’s useless unless it’s processed and unless it’s applied to something. Usually, that’s the big gap in a data strategy.
I like how you talk about how it’s like crude oil. You have to refine it and you have to do some work with it. Speak a little bit more about what might you have to do with some of your dirty data. Talk about that a little bit more.Analytics must follow the sequence of the five whys. Click To Tweet
What do you have to do with your dirty data? You’re touching the Holy Grail. We’re not even halfway through this show.
I go right at it. I don’t mess around.
You have to understand where that data comes from to fix the root cause because that data can always be corrected in the end. You can slap people on the wrist or you can publish big guide books on how to publish good data, but you’re only fixing a problem. You’re not addressing what’s at the root cause. What is that the root cause of the problem? Usually, people don’t know. People have no clue what good data mean and whatever they input is fine to them and they don’t care how it’s going to impact people after the downstream of them with analytics and stuff. It starts there. Second, there’s no incentive you put with data. It would be known if you could get a bonus. If the data you put in the systems are always right, you’d get a raise or get a percentage at the end of the year. It’s not happening.
You also have no time to pay attention to the data. It’s productivity. You get that in and nobody rewards or pays for the time that could be spent saying, “I’m going to make sure that there’s good data,” because people want short profits. What they don’t think about is, “We could get so much profit with that data, but we don’t want to invest the time to get it right.” You’ve got to work at this level to fix the root cause of data. That also requires a bit of culture that can be learned. You got to understand a bit of technology because you’re going to need tools to fix data, to be very efficient at fixing it. Unless you know these tools, it’s not going to work. Once again, Excel is not going to cut it and PowerPoint is not going to go to cut it on top of that.
The first thing I heard is if you’re starting to think about data and particularly for startups, you have to understand what you’re trying to get out of it. What is the data you’re looking for? It’s asking yourself, “What information do I need in the very beginning?” If you aren’t collecting that information, you don’t have the sources for the data, that’s your first barrier, right?
Yeah. For the startups I mentored or even I run, usually what is forgotten is, “What’s our business problem? What’s our business challenge?” Let’s not think about data first. Think about what value proposition we want to have in the end or what problem we want to solve in the end. Until we figure that out, there’s no point looking for data. We’re going to collect everything and anything and then God will come and recognize.
“If I collected all, maybe I can make sense of it and do something with it.”
Maybe you’d be lucky but unless you’re Google and you have tons of infrastructure to store all that data and then tons of PhDs and analysts to crunch the data. Even then, there’s probably a lot of ways this is not going to work. I work with startups like app developers and they say, “We’re collecting everything we can.” I ask them, “What do you want to do in the end?” “We want to sell that data.” “Who’s going to find that appealing?”
That’s not so in vogue right now.
That’s not exactly in vogue, but with marketing your data, you still need to fulfill a value proposition. Your data has to have some value and today if it’s just a laundry list of clients, that’s not going to help. You got to have data with a twist, with a spin and with some value and remain ethical, which brings a little bit of a barrier here because that’s why you need to know your data and know the technology or the techniques to handle it. That’s going to help you control anonymity and ethics. Going back to the key point, the business challenge or the value proposition you want in the end, unless you know what you want, you won’t be able to know which data you need to collect. You’re going to end collecting tons of information, many of it is useless. Because you collect a lot, the quality’s going to be bad. If you asked the right question and focused on the true elements that would have made the question relevant, you would have been able to control the quality, the intake, the ethics, the anonymity and differently answer the question that would bring value to your company.
This first question isn’t even about, “What data do I need to gather?” It’s going back to the heart of your startup. Why are you here? What is the problem that you’re trying to solve? It’s then thinking about what data I need to collect along the way in order to solve the problem.
I’m going to give you an example that applies to my own startup. People tell me, “You’ve got to do Google Analytics.” This is going to be very black and white but hear me on this, “Gauthier, you need to monitor every single page, Google Analytics and so forth.” I’m telling them, “I don’t sell on the web. I sell by referral and I sell when I deliver conferences. People come to me and buy my services. I don’t need that data.” “Everybody should be doing it.” No, my business data is not my Google Analytics. My business data is my ability to record my contacts and make sure I track the way I follow up with them and their interest. That’s it. If we still say, “Google analytics should monitor everything that happens on the website.” Let’s pretend this is relevant to me. Then the question is, “What do I want to do with that?” Hits on websites are useless. “We need to know if they are large clients and maybe if they already tried digital transformation and they failed, that’s why they would be needing me.” Then I know that in my business forum, these are the types of people I need to get.
Once again, monitoring Google Analytics won’t help. The right answer probably to get the data is if they come to my website, maybe ask them, “Have you ever run this?” Ask the question or maybe have a few links to download and say, “Here are the top twenty failures in digital transformation,” or “Here’s how to finally get your digital transformation projects successful.” If people download this, then I know they have an appetite for it. If they don’t, probably not. Thinking about the business problem leads me to the strategy, which leads me to the data and then I realized I don’t have that much to collect.
Even the systems or processes might be a CRM system or something that allows you to track or see more with interactions than it is something that’s purely monitoring your website. What you need to help you in that process, your technical solutions are going to be very different because of how your original business problem was defined. Yours being about monitoring and keeping up with the myriad of relationships at different statuses in your pipeline or in your processes and having very different and divergent needs. They’re all within the digital transformation space, but different needs within that. It’s even looking at the pipeline of people that come through and that’s people you meet at conferences and the referrals, which is a whole other set as well.Boil down data to the size, shape, and form that's going to work with the analytic recipe you're putting together. Click To Tweet
Those are going to come through email and phone calls and maybe the website, but through a whole variety of different channels. Where do you collect, monitor and look at that data? It’s an interesting way of thinking about the problem and saying, “I’m not an eCommerce website and I don’t need to look at how many shopping cart abandonments do I have. I don’t have a shopping cart. The analytics are different for me and unique based on what my problem is.” Looking back at, “What do I need to collect in order to make those actionable decisions?” is what I think I’m hearing you say. Is that a fair assessment?
That is right and don’t fall for all the buzz. I was a part of this startup and we had zero revenue. We never had any revenue, but focus number one was to get a humongous, ultra-expensive e-marketing tool. We don’t know how to sell. We didn’t even have the feedback from the market for having sold once, but yet we needed the heavy artillery because “everybody” has the heavy artillery in doing it. We’ve got to stop this. It’s not that everybody says everybody does it, that we should. We have a business problem to solve and that’s the only thing that should matter. How do we solve it?
How do we measure along the way? I’m a big fan of start somewhere and iterate. I know a lot of the reasons we’re philosophically aligned is you think the same way. You have people start somewhere and especially as a startup, you’re not going to create a grand, overarching analytic strategy. That’s way too much. You’ll have some strategy and some guidance, but if you were to recommend for a startup, once they get past some of the basics of, “I can answer why I’m here,” how would they take the next steps and thinking about what analytics they might need? Expand upon that a little bit more.
The type of analytics people will need should always start with basic questions. Usually, that’s a bit of a mistake people do. They say, “We want to do predictive right away.” You can’t predict until you know where you are and where you’re coming from and you know it well. Otherwise, you predict from a very blurry situation. It’s not bad to start with very simple measures of your business. Where do we sell? Who are our top five customers? We know that. Where do we sell? Is it always the same top five customers? You then move on. You keep on asking questions and from these very simple analytics that you get the hang of that you know your data is good. You start building more complex analytics which answers more questions.
Analytics must follow the sequence of the five whys where you wake up the first day and you say, “Why is our revenue stalling? It’s stalling because that product doesn’t work. Why doesn’t it work? It doesn’t work because we missed a few deals. Why did we miss a few deals?” The more why’s you ask, the harder to find the answer will be and this is where advanced analytics will have to kick in sometimes because your eyes and your senses won’t be able to guess anymore. They will have to predict or deduct or see the signal through the noise. You might want to say, “We want to go straight to that point because that’s what matters.” The problem is if it’s not built on top of data that have been profiled, cleaned, curated and understood, you won’t be able to go to the last why right away because you need all that data prior. If it’s not secured and mastered, you won’t be able to work.
It’s almost like you have to iterate through your whys. Why do we have an X, Y, Z situation? We better go collect a bunch of data to support that. Once we understand the why that’s happening there, we can ask ourselves the question again because it’s going to mean there’s a new set of information we would need at that point. If we sink through and structure and collect the information, we can build upon that. It’s not like the previous why is lost. That why is still relevant and still information I may want to track. I’m going to get more and more refined with the questions I ask and therefore, the information I need to collect.
Often people are coming to me and say, “I want to do predictive.” Do you have an immediate mapping of where you sold the past five years to get a sense of the lay of the land? They say, “Give us two weeks.” I’m barely joking. If you don’t have agility walking, you can imagine the agility you’re going to have flying and no, that doesn’t work.
For people that aren’t used to even thinking through the analytics, it can sound sexy to be like, “We were going to immediately go into predictive analytics and predicting where you’re going to sell and all these other things.” Talk about some of the baby steps of analytics, because I know I have a lot of audiences that might not be super technical and they’re on processes and approaches. Walk through the basics of the analytics chain and where people start and where they try to head over time.
The analytical chain is always the same. That’s what makes it somewhat easy to master because excellence comes with iterations and repetition because it’s the same process and if you do it each time, you’re going to get better at it. The process is the following. We start with the business question, “What do we need to solve?” Then there is a data collection step, which is we shall get that data and make that data available. This is where usually things break rapidly because if it takes you a month to get the data, forget the analytics because you’re going to have to run these analytics regularly and the effort is way too high. That phase is called data capture.
Then something that people usually hack in Excel and do a poor way at it or very inefficient way at it is you’re going to need to prep your data because no data comes cleaned and all organized. That’s called data modeling and data preparation. That’s like when you cook, you’re not going to cook with a whole truck of flour or a whole truck of sugar. You need little bags. Actually, you don’t need bags. Maybe you only need little tins and teaspoons and stuff. The same thing happens for data. You got to boil down data to the size and shape and form that’s going to work with the analytic recipe you’re putting together. You can then start analyzing the data. 80% of the work has been that collection and data prep. You can have to go through analyzing the data. If that prep phase has been done, it’s the easy way. This is where you put your business brain back on and you start thinking about, “Now, I need to find the answer to that question.” That’s pretty much it. You find the answer and you act upon the answer. Because if you find an answer you cannot act upon, you’ve achieved nothing and you start again.
That leads to a point where you’ve got to be pragmatic, “We found a great way to sell.” “How can we act upon it?” “We can’t. It would be too expensive.” “Why did you even look in that direction?” Everybody wants to be a strategic leader. This is where you have to think strategically and leading your people into rat holes by hacking data left and right only to figure out that you got an insight that you can’t act upon. That’s a waste of time. As a startup leader, you cannot afford this. You’ve got to see both ends of the tunnel, which is I got big business questions on one end I need to solve. I also know that to address these business questions, the answer has to be somewhat realistic or doable.
I’ve got to balance these two ends of the tunnel and between, “This is where the analytic change takes place, connecting a question to an answer, but an answer that will trigger an implementation or an action.” If that whole chain doesn’t hold, I’m sorry to be a bit of a party pooper, but you’re not achieving much. In a large corporation, you can please yourself for a while because everybody’s going to say, “It’s a great PowerPoint. It’s a great report, a great Tableau dashboard. Even though it’s useless as a startup person, you don’t have that luxury. You focus on things that are going to have any impact.
Thank you so much for coming on the show. I truly appreciate it. Where can people find out more about you if they want to connect to you or follow? You do such a variety of things. What’s the best way to find you?
You can find me on LinkedIn at Gauthier Vasseur and on Twitter, @GauthierVasseur. You can find my website, DatawiseAcademy.com, where I speak about what I do. I’m also a mentor at Google Launchpad. I love helping entrepreneurs and startup in that field. I can sell some dream, but my background is going mostly realistic approach to these things. I will always be very pleased to share it because that’s what I do now.
Thank you for reading. Don’t forget to rate me on iTunes or on Google podcasts. It’s time for you to go be wonderly.
- Gauthier Vasseur
- Gauthier Vasseur on LinkedIn
- @GauthierVasseur on Twitter
- Being Wonderly Podcast on iTunes
About Gauthier Vasseur
Trainer, professor, thought leader, and international speaker, Gauthier shares his cross-domain expertise with business teams and students to empower them with applicable techniques to master their data and analytics and become actors of their organization’s digital transformation.
10 years experience in leading universities, professional associations, and Fortune 500 companies and an engaging style bring impact, engagement, and content deeply ingrained in business reality. With an executive track record spans from large enterprises (Google, Oracle, Hyperion) to bootstrapped, series A and Pre-IPO companies (Semarchy, Trufa, TriNet). He ran many facets in Finance, Operation and Marketing, building up teams, analytics and scalable processes for growth and transparency.
A technology evangelist at heart, he has based his successes on a passion for innovation and their application for internal and external customer successes. An avid surfer, astronomer, and photographer he never misses an opportunity to push the boundaries of his knowledge to new limits.