This Insights Article from Mesh Digital LLC discusses how to make Data Analytics impactful. They use the analogy of a chef who prepares a meal but needs to understand the preferences of the diners and present the dish in a visually appealing way to make the food truly enjoyable.
Similarly, they argue that simply gathering data is not enough. They highlight the importance of identifying the data consumer's needs and preferences, understanding the competitive landscape, creating a collaborative process for data analysis, and designing a data product that can be easily integrated into workflows. They stress the necessity of adaptability in data products as data constantly evolves. Finally, they emphasize that the ultimate goal of data analytics is to leverage data insights to make meaningful and impactful decisions.
Mesh Digital LLC's Insights Full Article:
[00:00:00] Okay, so today let's really dive into this idea of data analytics, but like not just the surface level stuff, you know? We want to go deeper. We want to figure out how to make data like actually useful.
[00:00:14] Yeah.
[00:00:14] You know, how to take it from just like a bunch of numbers on a screen to something that gives us real insights that we can actually use.
[00:00:21] Yeah, yeah. No, it's a common problem, I think. You see a lot of companies, they've got tons of data, right? But they're almost like they're drowning in it.
[00:00:29] They're not quite sure how to actually get the knowledge out of it.
[00:00:32] It's like they've got all the ingredients for an amazing meal, but no recipe. Like how do you put it all together?
[00:00:37] Exactly.
[00:00:38] Luckily for us, we've got this paper from Mesh Digital LLC and it's like they're giving us the recipe right there like the master chefs of data.
[00:00:46] Yeah. It is a good analogy because it's, you're right, this paper does really lay out kind of a step-by-step process.
[00:00:52] And right away, the authors John Gugliotti and Michael D. Kleinberg, they start with this analogy about unused data.
[00:00:59] They compare it to a tree falling in the forest and nobody's around to hear it. Does it make a sound?
[00:01:03] Okay. I'm intrigued. I got to be honest. When you said data chefs, I was like, uh, I don't know about this, but I'm listening now.
[00:01:10] This is a good analogy. So how do we make sure our data is actually, you know, making a sound that we can hear?
[00:01:16] That's where this idea of the point of consumption comes in.
[00:01:19] Point of consumption.
[00:01:20] Yeah. So like, imagine you need to fix a leaky faucet, right? You don't want to read a whole textbook on plumbing just to find that one little fix.
[00:01:28] You want the answer right there when you need it.
[00:01:31] Yeah. Like a five minute YouTube video.
[00:01:33] Exactly. And that's how data should be too. You don't want to have to dig through a mountain of reports to get to that key insight.
[00:01:40] Okay. So it's about making data more accessible, more like right there when you need it.
[00:01:45] So how do we actually do that?
[00:01:46] Well, they have a whole process and the first step might surprise you.
[00:01:50] It's not about some fancy algorithm or anything. It's about really understanding who is going to be using this data.
[00:01:57] Who is the consumer of this data?
[00:01:59] Okay. So understanding your target audience makes sense.
[00:02:02] Yeah. Like what are their needs? What are their pain points? What are they actually trying to achieve with this data?
[00:02:07] And mesh digital, they're big proponents of going straight to the source.
[00:02:11] So like actually interviewing people, doing surveys, really digging deep into what people actually need.
[00:02:17] I like it. Get out of the office. Talk to some real people.
[00:02:20] Yeah. They had this example of this big retail company and they completely changed their whole inventory system just by like watching how their employees were using the old system.
[00:02:31] Just by observing.
[00:02:32] Yeah. And they were like, oh, wow, this is clearly not working. We need to fix this.
[00:02:36] Interesting. So we've we've put on our detective hats. We've gone out and we've interviewed our data consumers. What's next?
[00:02:42] So the next thing is what they call competitive intelligence.
[00:02:46] Oh, competitive intelligence. So like we're spying on the competition.
[00:02:50] Well, not exactly. OK.
[00:02:52] It's more about understanding what's working well for others in your industry and also what's not working.
[00:02:58] Like what are the mistakes that other people have made that you can avoid?
[00:03:02] You know, learn from their mistakes. Don't reinvent the wheel.
[00:03:05] Exactly. They actually gave this really interesting example of a financial services company and they actually learned a lot from a video game company.
[00:03:13] Really?
[00:03:13] Yeah. Like how they were using data to get people more engaged with their games.
[00:03:17] Wow. That's fascinating. See, this is why I love these deep dives.
[00:03:21] You make these connections that you never would have thought of before.
[00:03:24] So we've we've done our research. We've scoped out the competition.
[00:03:27] We're feeling like dated detectives over here.
[00:03:30] What's the next step in our data journey?
[00:03:33] OK, so now's when we actually get into like building the data product and mesh digital.
[00:03:38] They really stress this collaborative approach.
[00:03:42] So you've got your data scientists, obviously, but then you've also got data engineers.
[00:03:46] And then this is interesting.
[00:03:48] They also bring in visualization experts.
[00:03:50] Visualization experts. I wouldn't have thought of that.
[00:03:52] So it's not just about the numbers themselves.
[00:03:54] It's about how you present them.
[00:03:55] Exactly. Like you could have the most amazing data analysis in the world.
[00:03:58] But if you present it in a way that's boring or confusing.
[00:04:03] Nobody's going to look at it.
[00:04:04] Exactly. And that's where these visualization experts come in.
[00:04:08] They can help make that data come alive.
[00:04:10] They can make it visually appealing and easy to understand.
[00:04:13] Yeah. It's like you need the chef to make the meal taste good, but then you need somebody to actually plate it, you know, to make it look appetizing.
[00:04:23] Exactly. It's the presentation. It's everything. Right.
[00:04:25] And what's cool is mesh digital. They don't stop there.
[00:04:28] They even talk about like agile methodologies and DevSecOps.
[00:04:32] I don't know if you're familiar with those terms.
[00:04:34] I've heard of them, but they're not something I usually think about when it comes to data.
[00:04:38] Right. You think like data that's like the IT department or something.
[00:04:41] Right. Right.
[00:04:42] But they're saying like, no, this is really important.
[00:04:44] If you want to build a data product that's actually going to last, that's going to be secure, that's going to be adaptable, you need to be thinking about these things.
[00:04:52] So there's that word again, adaptable. Why is that so important when we're talking about data?
[00:04:57] Well, because data itself is always changing, right?
[00:05:00] Yeah, that's true. Like what's true today might not be true tomorrow.
[00:05:03] Like consumer behavior changes all the time. Markets go up and down.
[00:05:07] Right.
[00:05:08] So if your data product is not able to adapt to those changes, it's going to become obsolete really quickly.
[00:05:14] It's like using a map from like five years ago to try to get around.
[00:05:18] You're going to run into some problems.
[00:05:19] Exactly. And they had this example in the paper of this social media company and they had this data product that was working really well for them.
[00:05:27] And then like their whole user base changed like almost overnight.
[00:05:31] Oh, no.
[00:05:32] And they didn't adapt their data product to account for that.
[00:05:35] And so suddenly all their insights were wrong and they were making bad decisions based on that bad data.
[00:05:41] So how do you prevent that? How do you keep your data product from becoming like outdated like that?
[00:05:47] Well, they talk a lot about the importance of evaluating and retraining your models.
[00:05:51] And I think that's something that people don't always think about.
[00:05:53] Like you build the model and then you think, OK, I'm done.
[00:05:55] Right. Set it and forget it.
[00:05:57] Exactly. But that's not how it works.
[00:05:59] You've got to you've got to keep training it like an athlete.
[00:06:01] You know, an athlete can't just like rest on their laurels.
[00:06:04] They have to keep training.
[00:06:05] Data models need their protein shakes and their treadmills, I guess.
[00:06:09] Exactly. And so Mesh Digital, they're big proponents of using these tools like observable notebooks.
[00:06:15] And these tools basically allow data scientists to kind of like tinker with the models.
[00:06:19] OK.
[00:06:19] They can try different things.
[00:06:21] They can see what's working, what's not working.
[00:06:23] They can visualize the data in different ways.
[00:06:25] It's like this really cool iterative process.
[00:06:27] So they're constantly refining and improving the model.
[00:06:30] So it's like a digital art studio.
[00:06:32] You know, they can they can try different things and see what looks best.
[00:06:35] Yeah. And once they've got something, they like they can just like take that code and plug it right into their application.
[00:06:41] So it's very seamless.
[00:06:42] That's cool. But, you know, even with all of this, even if we have this amazing data product that's beautiful and adaptable and it gives us all these insights,
[00:06:51] it's not going to do us any good if people don't actually use it right.
[00:06:53] Exactly. A hundred percent.
[00:06:54] I mean, you can lead a horse to water.
[00:06:56] Right. You can have the best data in the world.
[00:06:59] But you can't make them drink it.
[00:07:00] Exactly.
[00:07:01] So Mesh Digital, they really emphasize this idea of like integrating the data into a workflow.
[00:07:07] OK.
[00:07:08] Like meet people where they are so they don't have to go out of their way to use this data.
[00:07:12] It's just like right there in the tools that they're already using.
[00:07:15] So like seamlessly integrating it.
[00:07:17] Exactly.
[00:07:17] And they gave this great example of the sales team that was just like drowning in leads.
[00:07:22] OK. So they were drowning in leads.
[00:07:24] Did they even know who to call?
[00:07:25] No. It was like they had the CRM system, but it was really clunky.
[00:07:30] Yeah. I've been there.
[00:07:31] And they were just like overwhelmed with data, but they weren't getting the insights they needed from it.
[00:07:36] Right. So it's like having a library full of books, but no Dewey Decimal system.
[00:07:39] Exactly. They had all this information, but they didn't know what to do with it.
[00:07:43] Right.
[00:07:43] So what they ended up doing was they integrated this data driven lead scoring system right into their workflow.
[00:07:51] OK.
[00:07:51] So basically it was telling them like these are the hottest leads focus on these first.
[00:07:56] Oh, that would be amazing.
[00:07:58] Like it's telling you who to call, when to call them, what to say.
[00:08:00] Yeah. It's like having a superpower, you know, and their sales just went through the roof after that.
[00:08:05] Because they were finally working smarter, not harder.
[00:08:07] Exactly. And that's what I love about this paper. It's all about making data work for you, not the other way around.
[00:08:13] Right. Make it your superpower instead of your kryptonite.
[00:08:17] Exactly.
[00:08:17] I love it.
[00:08:18] Well, we've covered a lot of ground here today. Everything from understanding your data consumer to scoping out the competition to building these really adaptable tech stacks to making sure that you integrate the data into your workflow.
[00:08:32] It's a lot.
[00:08:32] It is a lot. But I think it's all really important stuff if you want to make sure that your data is actually useful.
[00:08:38] Definitely.
[00:08:38] So if you had to boil it all down to one key takeaway, what would it be?
[00:08:42] I think it's that data analytics is only as good as the impact it has.
[00:08:47] Okay. I like that data with impact.
[00:08:49] Yeah. It's not about just like having the coolest algorithms or the most data. It's about using that data to actually make a difference.
[00:08:57] Data with a purpose.
[00:08:59] Well said.
[00:08:59] So as we wrap up here, I want to leave our listeners with something to think about.
[00:09:03] Think about a decision you make on a regular basis.
[00:09:06] Maybe it's at work. Maybe it's in your personal life.
[00:09:09] How could data delivered at the right time in the right way help you make that decision even better?
[00:09:15] Oh, it's a good one.
[00:09:16] Yeah. What insights are you missing out on just because the data isn't there or it's not presented in a way that makes sense to you?
[00:09:22] Yeah. Food for thought for sure.
[00:09:24] Absolutely.
[00:09:25] Well, this has been a fascinating deep dive. I've learned a lot.
[00:09:28] I hope you have too.
[00:09:29] If you're ready to take your data from zero to hero, be sure to check out this paper from Mesh Digital LLC.
[00:09:37] And until next time, happy analyzing.