WEBVTT
00:00:00.100 --> 00:00:01.082
Hey, what is up?
00:00:01.082 --> 00:00:04.312
Welcome to this episode of the Wantrepreneur to Entrepreneur podcast.
00:00:04.312 --> 00:00:34.430
As always, I'm your host, brian LoFermento, and holy cow, I know I'm excited for every single episode, but today's guest already even just his energy gets me so excited, because this is someone who not only are we all going to be wowed at what he's built and put into the world and the wonderful world of data and analytics to actually drive us forward, but this is someone who I'm convinced his energy is equally excitable as mine is when it comes to just all things entrepreneurship and, even broader than that, all things possibilities.
00:00:34.430 --> 00:00:36.207
This is a dreamer and a doer.
00:00:36.207 --> 00:00:38.226
So let me introduce you to today's guest.
00:00:38.226 --> 00:00:39.706
His name is Arjun Badesi.
00:00:40.100 --> 00:00:42.179
Arjun is a visionary analytics leader.
00:00:42.179 --> 00:00:46.691
He's a tireless problem solver and the entrepreneurial mind behind Orkanaai.
00:00:46.691 --> 00:00:50.429
Wait until you hear about what he's built and put into the world.
00:00:50.429 --> 00:01:05.328
With over 15 years of experience transforming data into powerful business strategies, arjun has helped companies unlock over $600 million in incremental revenue by rethinking how data drives decisions.
00:01:05.328 --> 00:01:24.668
His journey spans industries like telecom, saas and pharmaceuticals, where he's led teams to solve high-stakes challenges, from predicting customer churn to optimizing over $300 million in marketing spend, with a master's in operations, research from UC Berkeley and an MBA from one of India's top schools.
00:01:24.668 --> 00:01:29.147
Arjun combines technical brilliance with a deep understanding of business needs.
00:01:29.448 --> 00:01:39.409
Now this is going to be a fun one, because I am so excited personally to hear more about all the wonderful ways that his company and his tool that he's created and launched into the world.
00:01:39.409 --> 00:01:51.992
Orkanaai is changing the game, because their mission is to make analytics effortless and impactful for business leaders, turning complex data into clear, actionable insights that drive growth.
00:01:51.992 --> 00:01:53.828
So I think that's a perfect primer.
00:01:53.828 --> 00:01:55.525
I'm not going to say anything else.
00:01:55.525 --> 00:01:58.549
Let's dive straight into my interview with Arjun Bidesi.
00:01:58.549 --> 00:02:05.150
All right, arjun, I am so very excited that you're here with us today.
00:02:05.150 --> 00:02:06.123
First things first.
00:02:06.123 --> 00:02:08.825
Welcome to the show, thank you so much, Brian.
00:02:09.200 --> 00:02:10.305
That was a great introduction.
00:02:10.305 --> 00:02:11.087
Thank you so much.
00:02:11.560 --> 00:02:15.504
Well, Arjun, you make it easy because truly, I mean it you are excitable.
00:02:15.504 --> 00:02:17.163
All the things I found about you.
00:02:17.163 --> 00:02:18.187
I've read about you.
00:02:18.187 --> 00:02:19.610
Your energy shines through.
00:02:19.610 --> 00:02:20.925
So let's kick it off there.
00:02:20.925 --> 00:02:22.824
Take us beyond the bio who's, Arjun.
00:02:22.824 --> 00:02:22.905
How?
00:02:22.905 --> 00:02:23.546
Let's kick it off there.
00:02:23.546 --> 00:02:24.389
Take us beyond the bio who's.
00:02:26.033 --> 00:02:26.194
Arjun.
00:02:26.194 --> 00:02:27.197
How'd you start doing all these cool things?
00:02:27.197 --> 00:02:27.679
Absolutely, brian.
00:02:27.679 --> 00:02:32.792
I've been someone who's been pulled into at least 10,000 fire drills in my career.
00:02:32.792 --> 00:02:44.733
Now a fire drill happens when, for example, a CFO of a high growth company wakes up at 6 am on a Sunday and notices that last week's profits declined by about 15% week on week.
00:02:44.733 --> 00:02:48.344
Guess what happens during the course of the rest of the week in that company.
00:02:48.344 --> 00:02:49.769
It's chaos.
00:02:49.769 --> 00:03:01.026
Revenue team, data teams, marketing teams, growth teams, customer success teams almost all aspects that inform product margin are running their investigations.
00:03:01.026 --> 00:03:04.491
The kicker it's extremely high pressure.
00:03:04.491 --> 00:03:07.921
Everybody that reports to the CFO is on the hook.
00:03:07.921 --> 00:03:12.411
Everybody CFO reports to CEO is set board are on the hook.
00:03:12.411 --> 00:03:23.012
Now some data scientists are on a vacation, some business leaders are on a way and everything in this place looks as investigations are going on in silo.
00:03:24.682 --> 00:03:40.973
It used to keep reminding me of this Indian parable that I believe was eventually translated by John Sachs, on six blind men and the elephant, where everyone, while discovering the elephant, is partly in the right, but all are in the wrong.
00:03:40.973 --> 00:03:51.669
Sometimes the elephant disappears, sometimes a bigger elephant shows itself, and very rarely the entire picture or the actual root cause is uncovered.
00:03:51.669 --> 00:03:53.995
For over 15 years.
00:03:53.995 --> 00:04:08.861
I had seen this across several industries SaaS, pharma, marketplaces and all company sizes, be it a series B company, a series G company or a billion-dollar or multi-billion-dollar companies.
00:04:08.861 --> 00:04:12.792
They're always wanting to know what happened, how did it happen, how do we fix this?
00:04:12.792 --> 00:04:17.028
And most questions, at best, were left unanswered.
00:04:17.028 --> 00:04:32.673
I got to witness how disconnected metrics, rigid dashboards and overwhelmed data teams which I have been a huge part of mostly have been functioning in silos and leaving the leaders flying semi-blind, if not fully blind.
00:04:33.821 --> 00:04:47.485
Over the course, the intelligent laziness in me tried several ways to ease my life, came up with these few accelerators code snippets, excel workbooks so that I could at least think of ways to get to the answer faster.
00:04:47.485 --> 00:04:53.351
Then I started to realize, oh, there is probably a way for me to productize this.
00:04:53.351 --> 00:04:56.187
That's what inspired me to build Orkhana AI.
00:04:56.187 --> 00:05:06.718
It's a product that works like a super analyst and can be used as the first line of defense or first thing to get to.
00:05:06.718 --> 00:05:12.646
If you have a question, be it business leaders or data scientists, you can start your hypothesis here.
00:05:12.646 --> 00:05:18.411
It's built to deliver answers explainably, reliably and always accurately.
00:05:18.411 --> 00:05:25.173
And now, thanks to LLMs in GenAI, it does the translation of these insights in plain English.
00:05:25.173 --> 00:05:31.213
Net-net Orkana empowers leaders to stop guessing and start winning.
00:05:31.213 --> 00:05:33.783
The goal is to reduce chaos.
00:05:33.783 --> 00:05:38.074
Convert that into clarity every single time.
00:05:39.040 --> 00:05:48.411
Yes, I love that overview, arjun, especially because it fits in so nicely with your backstory of putting so many fires out in your career and talk about organizing the chaos.
00:05:48.411 --> 00:05:50.427
Arjun, I'm going to publicly call this out.
00:05:50.427 --> 00:06:02.596
I've been so excited to have this conversation because typically we don't go this early on into a guest's business, but I think that your business so clearly illustrates not only the problem, but for sure it is the solution.
00:06:02.596 --> 00:06:15.911
In so many ways, I'm personally excited to go deeper into this world because here's what I really want to publicly praise you guys for is that I've never seen a solution so clearly paint the image of the gap that it fills.
00:06:16.232 --> 00:06:24.507
So on your website, as soon as people get to your website, they can see a real life use case of asking Orkana, why did revenue drop last week?
00:06:24.507 --> 00:06:27.302
And that's a real, tangible business question.
00:06:27.302 --> 00:06:32.062
We don't have to understand data, we don't have to understand AI, we don't have to understand analytics.
00:06:32.062 --> 00:06:37.721
We all understand in English how to ask why did revenue drop last week?
00:06:37.721 --> 00:06:39.925
Walk us through that use case.
00:06:39.925 --> 00:06:41.389
Let's roll with that for a little bit.
00:06:41.389 --> 00:06:50.867
Walk us through how Orkana, or data, or AI or all of the above, makes sense of these real life problems to give real life insights and solutions.
00:06:52.079 --> 00:06:52.944
Absolutely, brian.
00:06:52.944 --> 00:07:08.995
It dabbles into a little bit of what I have gathered over the last 15 years why people look at metrics or their business in the orgs they own.
00:07:08.995 --> 00:07:18.846
With so much of experience across so many industries, I've started to look at companies as a machine, a machine that has a whole lot of gears.
00:07:18.846 --> 00:07:22.882
Each of these gears is a metric of its own and they're interrelated.
00:07:22.882 --> 00:07:30.709
A revenue metric is impacted by a churn metric, also impacted by a spend metric is impacted by engagement metrics, so on and so forth.
00:07:30.709 --> 00:07:48.601
So I have now started to really intuitively build models that can trickle down a complete business or an org into organization, into a whole lot of these metrics that function within each other's dependencies.
00:07:48.601 --> 00:08:04.425
The best part about this setting this up as a framework or a web-like feature map for a given company is it's going to function in a very thorough and robust manner when a question comes to mind.
00:08:04.930 --> 00:08:09.170
Because when you say, why did your revenue decrease, there may be so many reasons for it.
00:08:09.170 --> 00:08:13.545
Maybe you spent a little less, maybe the revenue met data is incorrect.
00:08:13.545 --> 00:08:21.562
There are yet a bunch of affiliate channels that are supposed to convert and share their revenue or share those numbers as to how they look like those numbers as to how they look like.
00:08:21.562 --> 00:08:25.725
Maybe there was a storm somewhere, maybe this was.
00:08:25.725 --> 00:08:34.572
This happens every Monday of the first Monday of the month, maybe if there is a holiday, like a Labor Day weekend, et cetera.
00:08:34.572 --> 00:08:46.081
So there are so many things that happen and trying to uncover this if left to people, especially under high pressure, you want to look at the first thing.
00:08:46.081 --> 00:08:51.602
That kind of shows the answer and says you know what I started to see, this is what it may look like and we just probably start solving for it.
00:08:51.602 --> 00:09:05.057
But with Autana, it does this and, with all of explainability, 100% of the times it doesn't leave any stone unturned.
00:09:05.057 --> 00:09:06.422
It says all right, revenue decrease.
00:09:06.701 --> 00:09:07.323
Did spend change?
00:09:07.323 --> 00:09:10.611
If yes, how much Did revenue come in higher or lower?
00:09:10.611 --> 00:09:15.437
Are all the data sources that come in for revenue change the most Come in accurately?
00:09:15.437 --> 00:09:16.863
Did some of them change?
00:09:16.863 --> 00:09:21.235
So it tries to understand right from saying did revenue change?
00:09:21.235 --> 00:09:25.931
If revenue change, was it from the current customer or from the new customers?
00:09:26.561 --> 00:09:32.919
If it was from the new customers, were they because of acquisitions that happened fewer or conversions that happened fewer?
00:09:32.919 --> 00:09:37.875
Now, if acquisitions were fewer, did meta bring in fewer leads?
00:09:37.875 --> 00:09:41.167
Or Facebook or TikTok bring in fewer leads, so on and so forth.
00:09:41.167 --> 00:09:53.461
So, given this metric map and the way we have set up the analytics engine, which is like the proprietary engine for our product, it just helps you uncover with this simple questions.
00:09:53.461 --> 00:10:05.552
It does all of this math like a very diligent, super fast analyst, brings up that answer to say here is what are the three things that seem to have changed the most?
00:10:05.552 --> 00:10:06.192
Can we?
00:10:06.192 --> 00:10:12.403
Do you want to double click into one of these and uncover what could be the next level of root cause that that you'd want to look?
00:10:12.403 --> 00:10:23.870
That's how it kind of really helps expedite from questions to insights, uh, in a very reliable and explainable manner yes, gosh, gosh, arjun.
00:10:23.971 --> 00:10:25.714
So much to unpack from that.
00:10:25.714 --> 00:10:26.966
Obviously, it's very powerful.
00:10:26.966 --> 00:10:33.808
What I'm thinking as I'm hearing you talk about this is I'm thinking well, we all know that big businesses have data departments.
00:10:33.808 --> 00:10:55.092
A lot of us, our first hire might be as a virtual assistant or a marketing department, but the truth is we now have an analytics department at our fingertips through what you've built with Orkana, and that's so cool to be able to onboard an employee, so to speak, that is powered by technology, powered by AI, powered by all the data things that you just shared with us.
00:10:55.480 --> 00:11:01.244
So my first question is when I hear you talk about these different data sources I want to unleash in all of my businesses, arjun.
00:11:01.244 --> 00:11:05.731
So you and I are going to talk about this off the air, but with that in mind, where does it live?
00:11:05.731 --> 00:11:07.975
How does it obtain all of these different things?
00:11:07.975 --> 00:11:20.101
Because I'll just transparently tell you this I was a math tutor in college and I love regression analysis, because you can have as many variables as you want and there's no way the human mind can make sense of multivariate regressions.
00:11:20.101 --> 00:11:24.469
But that's what technology and especially now we live in the AI era that's what it can do.
00:11:24.469 --> 00:11:28.774
I don't know how to make sense of all these different components of my businesses.
00:11:28.774 --> 00:11:31.524
You listed social media, you talk about paid ads.
00:11:31.524 --> 00:11:33.086
You talk about traffic to our website.
00:11:33.086 --> 00:11:34.426
There's so many different things.
00:11:34.426 --> 00:11:35.884
Where does Arcana live?
00:11:35.884 --> 00:11:38.268
How does it plug into all of these data sources?
00:11:38.268 --> 00:11:41.508
How do we get it to have that overview of our entire business?
00:11:43.091 --> 00:11:43.673
Absolutely.
00:11:43.673 --> 00:11:49.399
It just needs a pipeline to pull in this data.
00:11:49.399 --> 00:12:02.565
It either pulls it in directly from the Meta's API, tiktok's API, stripe's API for any revenue information or billing information, so on and so forth, or it can connect to your data warehouse as it sits today.
00:12:02.565 --> 00:12:14.268
Because a whole lot of these reports that big, large companies, for example, have that have these marketing dashboards or this revenue operations dashboards, have a reporting layer where all of these metrics are kind of defined.
00:12:14.268 --> 00:12:25.186
But business users are not really well trained to go beyond the reporting layer to say, hey, can I get a double click and why did this change?
00:12:25.186 --> 00:12:26.109
So on, so forth.
00:12:26.109 --> 00:12:45.368
So the data that sits is sitting wherever this is the, the data is as readily available as possible and between wherever that data is to what is needed to make sense of a business question, we start slow.
00:12:45.368 --> 00:13:04.057
When we onboard Arcana into any company, we identify a department which is probably the most underserved or the most insights hungry, because it changes a lot and that department's performance is probably very critical for the entire strategic goals of that company to come through.
00:13:04.057 --> 00:13:07.355
So what we do is we identify a use case.
00:13:07.355 --> 00:13:12.577
We identify what are the two or three metrics that really need to be nailed within this organization.
00:13:12.577 --> 00:13:26.499
We identify and come up with all the dependency maps for those two to three metrics to ensure that if profit margin or contribution margin is what you need to obsess about, if you want to obsess about higher lifetime value, that's yet another metric.
00:13:26.499 --> 00:13:35.111
And then we identify work backwards from there to keep sourcing these data sets or ways to identify what these metrics sit like.
00:13:35.965 --> 00:13:53.071
So we have a logic engine in place, because LLMs surely can do the natural language understanding, but there needs to be a well-defined and reproducible logic that then goes and keeps these analytics very quick and robust.
00:13:53.071 --> 00:13:54.854
So we start from a use case.
00:13:54.854 --> 00:13:58.788
We identify what are these metrics that seem like dependencies.
00:13:58.788 --> 00:14:06.652
We test the levels of dependencies, identify what's the best, most processed data source If processed, great If not.
00:14:06.652 --> 00:14:13.591
We directly hook up to the APIs of the world and then the process is up and running.
00:14:13.591 --> 00:14:21.394
Think about it as a process which makes you emulate a finance analyst or a marketing analyst, a growth analyst.
00:14:21.394 --> 00:14:22.537
And that's how we start.
00:14:22.537 --> 00:14:32.000
We start with picking one department, picking a few set of use cases where we think they are the most underserved and the companies like, really want to solve for that, and then that's it.
00:14:32.000 --> 00:14:36.331
We just land and expand based on you proving our use case and value.
00:14:37.113 --> 00:14:49.831
Arjun, I love the way you articulate that we land and expand and I think that that expansive thinking, it's on full display for us here today, because you just introduced a term, arjun I've never heard those words combined a dependency map.
00:14:49.831 --> 00:15:08.972
You pick a metric, but then and you and I can obviously see each other A lot of listeners can't see us right now, but you showed that trickle down effect of the fact that that metric is driven by probably 20 to 50 other metrics along the way Paint that picture for us and let's use revenue as the example, because I know a lot of people when we talk about revenue, they just go straight to sales.
00:15:08.972 --> 00:15:15.491
You know, am I converting my sales conversations Give us a real life overview into some of that dependency map that you painted.
00:15:15.491 --> 00:15:16.293
That picture for us.
00:15:17.416 --> 00:15:18.317
Absolutely, brian.
00:15:18.317 --> 00:15:19.038
I love this.
00:15:19.038 --> 00:15:21.173
I live and breathe this.
00:15:21.173 --> 00:15:32.111
I see so much of excitement when I translate a metric like revenue into a meaningful set of things that business users can actually control and really diagnose and improve.
00:15:32.111 --> 00:15:35.846
So revenue is a function of say two things.
00:15:35.846 --> 00:15:41.888
One is the number of customers and the revenue being generated from each of these customers.
00:15:41.888 --> 00:15:50.913
Now, if revenue changed, you could see look at these two and say did the number of customers go down or is the average revenue per customer go down?
00:15:50.913 --> 00:16:02.019
Now, if assume we go with the average revenue per customer, we then go down to say, of these average revenue per customers, are there certain segments or some certain categories of these average revenue per customers?
00:16:02.019 --> 00:16:09.034
Are there certain segments or some certain categories of these customers that have suddenly started to give us less revenue, generate less revenue?
00:16:09.034 --> 00:16:17.049
Or are there any new versus existing customers that are signing up for fewer, lesser tier values or lesser tier services?
00:16:17.049 --> 00:16:28.355
And that's how we try and understand is revenue per customer the key thing that we need to double click into and if yes, which segment or which kind of an offering?
00:16:28.355 --> 00:16:33.056
And then from there we get down into saying were they on the trial period for shorter period?
00:16:33.056 --> 00:16:35.173
What industries are they coming in from?
00:16:35.173 --> 00:16:39.135
Is there something else that we can fill in to see why this average is going down?
00:16:39.504 --> 00:16:44.606
This is like one area or one path that you can take when you're talking about revenue.
00:16:44.606 --> 00:16:48.296
Similarly, you check out the other path, which is customers.
00:16:48.296 --> 00:16:50.006
Now the customers decrease.
00:16:50.006 --> 00:16:55.519
If yes, did more people churn or did fewer new ones come in?
00:16:55.519 --> 00:17:02.835
And if the fewer new ones come in, did they come in from inbound sales, outbound sales?
00:17:03.615 --> 00:17:19.128
Were there any other aspects about these customers that really ensured that the dip happened for a particular industry or a particular geography where customers were usually coming in at quite a healthy rate, but suddenly you don't see something?
00:17:20.050 --> 00:17:33.637
So that way you're able to like break revenue into two parts Now break customers into another two, three, four parts, break each of these into a few more parts and you keep going as deeper as possible, as long as data allows.
00:17:34.464 --> 00:17:59.314
And that's where we say you give us the most granular data that you want and you will exactly be able to pinpoint, to say you know what this campaign at Meta, where we gave a $100 voucher, brought in a lot of customers, but because the value was so much, even the low intent ones came in used up that $100, but never converted, so your revenue did not really show that healthy metric.
00:17:59.314 --> 00:18:11.416
Eventually, because that campaign of you using a $100 voucher was not the profit maximizing one, and this usually would have taken a week, several weeks or sometimes.
00:18:11.416 --> 00:18:32.271
This never gets uncovered, but because this map exists and we know Orkana is always going through each of these details to uncover what moved the most, it brings this up with a series of questions so that you then know there is something on the meta campaign side that you need to fix the next time you roll out a promotional offer, for example.
00:18:32.772 --> 00:18:34.477
Yeah, arjun, so cool.
00:18:34.477 --> 00:18:46.644
What I think is really fascinating about the way that you talk about data is, I feel like again I'm going to come back to that expansive thinking that you have, and I know that for you personally, it comes through the fact that you've worked across so many different industries.
00:18:46.644 --> 00:18:48.953
You've seen behind the scenes of a lot of different companies.
00:18:48.953 --> 00:18:59.092
I can tell you from my vantage point just being an entrepreneur for 16 years and obviously I talk to entrepreneurs every single day of my life I find that certain industries are more data friendly.
00:18:59.092 --> 00:19:03.239
For example, anytime I talk to someone in e-commerce, they love data.
00:19:03.239 --> 00:19:05.450
They get as many data points as they possibly can.
00:19:05.509 --> 00:19:19.731
But, arjun, I'm going to let you speak directly to the listeners here in the audience today is that a lot of times, people outside of that, you know, in the service-based business, for example I'll pick on them for a minute They've convinced themselves that, oh, I'm in the people business.
00:19:19.731 --> 00:19:26.553
You know mine isn't as data heavy and I don't have conversion analytics from my website and I don't have these long marketing files.
00:19:26.553 --> 00:19:31.575
What's your response to those people who think that their industry or business model is different?
00:19:35.287 --> 00:19:39.153
That's a great question, brian, and this is something that we sometimes dabble with.
00:19:39.153 --> 00:19:48.853
Thankfully, this has become less of a problem off late because people have gotten used to, hey, it's important data wise.
00:19:48.853 --> 00:19:50.196
But this is how I explain it.
00:19:50.196 --> 00:19:54.699
Right, my first thing is you cannot improve something you can't measure.
00:19:54.699 --> 00:20:01.926
Now, if you're in the world of business, you're there to feed mouths, you're there to create an impact, you're there to generate revenue.
00:20:01.926 --> 00:20:15.487
Now, if you're having a business that says I don't want to increase my revenue, I'm okay if I keep declining my sales, I think that's an outlier.
00:20:15.487 --> 00:20:43.576
But in most cases, when you say you know what, if you understand what's driving your revenue, you can figure out ways to double down on what's working, reduce the aspects about what's harming you and naturally, you can probably if not grow automatically serve so much better to your customers, serve so much better to your team members by giving them whatever advantages or translating those profits across the world whatever advantages or translating those profits across the world.
00:20:43.576 --> 00:20:50.188
So that's how I try and bring this up, saying everything can be measured.
00:20:50.188 --> 00:21:02.750
And once you learn how things can be measured, you can identify what drives these metrics and then you can figure out ways to improve that and that's how I convinced them, saying you know what?
00:21:02.750 --> 00:21:06.238
Give me an example, tell me what's your biggest pain point.
00:21:06.238 --> 00:21:10.153
And they're like customers churn but there's basically like 25 things that happen.
00:21:10.153 --> 00:21:13.112
One of them is it's very disparate, etc.
00:21:13.112 --> 00:21:14.336
And I'm like sure, give us your data.
00:21:14.336 --> 00:21:19.990
If you're running this business for three years, you have about 200 customers that probably churn.
00:21:19.990 --> 00:21:21.632
We are going to find patterns.
00:21:21.632 --> 00:21:34.279
We're going to find patterns that lead to churn events, because that churn happens like way before you probably start reducing your product engagement, you start to down deal your service.
00:21:34.279 --> 00:21:38.070
We'll identify what's really happening here and we can help.
00:21:38.070 --> 00:21:38.732
You know that.
00:21:38.732 --> 00:21:39.193
You know what.
00:21:39.193 --> 00:21:43.614
There is something that is going to happen with these set of customers.
00:21:43.614 --> 00:21:47.393
If you don't want to lose revenue, you might as well just act upon it.
00:21:47.393 --> 00:21:57.634
And that really helps them understand that you're not really knowing what's happening or when is it going to happen, but how to counteract it.
00:21:57.634 --> 00:22:00.574
And if they ask me saying hey, how do you counteract it?
00:22:00.574 --> 00:22:05.636
Because that's also yet another very obvious question that come in saying how do you even know what things are working?
00:22:05.636 --> 00:22:07.627
We like that's data.
00:22:08.028 --> 00:22:10.994
We look at similar customers who are not churning.
00:22:10.994 --> 00:22:13.426
We're going to see what are they using?
00:22:13.426 --> 00:22:20.653
What are the various features about that product, about your product or service that they're like really engaging with?
00:22:20.653 --> 00:22:29.074
And then you're going to recommend saying you know what they are on the words of churning, but these ones are still very loyal because they're using these features, which they are not.
00:22:29.074 --> 00:22:31.065
Why don't you try and promote those features?
00:22:31.486 --> 00:22:35.430
So now you don't just know that, hey, churn is like very unpredictable.
00:22:35.430 --> 00:22:36.835
You know churn is predictable.
00:22:36.835 --> 00:22:39.750
Now you know what makes it predictable.
00:22:39.750 --> 00:23:04.516
And then you also know how to reverse that churn prediction and and then it just light, lights up a bulb and they're like let's get this started and and this is a very live example, because we're doing this for a series b uh tech company that having a huge problem with churn and we were like let's fix this for you.
00:23:04.516 --> 00:23:11.926
We're going to make it so automatic that the partner success teams will get just like alert saying risk of churn, this is why they're churning.
00:23:11.926 --> 00:23:22.777
This is what you could do to get this right comes on your slack as an and you have your to-do for that day or week yeah, arjun, hearing you talk about this it's very important, I think, in our conversation today to talk about this.
00:23:22.696 --> 00:23:28.218
It's very important, I think, in our conversation today to talk about the fact that we're not talking about data to make sense of the past.
00:23:28.218 --> 00:23:35.771
What I'm really hearing and the fact that you come with real life examples is so cool, because what I'm hearing is that we're using the past.
00:23:35.771 --> 00:23:43.948
We're using data points that we can get our hands on, but we're using it to look towards the future and to make smarter strategic decisions and to change our actions.
00:23:43.948 --> 00:24:01.071
Cross that bridge for us, for listeners who have never operated this way and it's so cool to hear that, with a Series B company, you're the one who's the catalyst for them to start operating this way For people who haven't used data in that way to make forward looking decisions we all do it at tax time, ar Arjun.
00:24:01.092 --> 00:24:08.074
We can't help but look at our old data when it comes to taxes, but talk to us about using it to make forward-looking decisions.
00:24:08.074 --> 00:24:16.250
What are examples of the different decisions that we can take and that we can make in our businesses thanks to all this work that you're doing and that Orkana can pump out for us?
00:24:17.953 --> 00:24:18.796
absolutely, brian.
00:24:18.796 --> 00:24:24.718
So the entire gamut of analytics can be bucketed into four key aspects.
00:24:24.718 --> 00:24:28.049
One is diagnostic, sorry.
00:24:28.049 --> 00:24:35.367
One is descriptive, which basically says just explain what really has happened, what happened, where did it happen.
00:24:35.367 --> 00:24:38.453
Then there's diagnostic, which is why did it happen?
00:24:38.453 --> 00:24:40.198
Why did it happen, what caused it?
00:24:40.198 --> 00:24:50.659
Then you move from that to predictive, which is is there a way for you to?
00:24:50.659 --> 00:24:54.673
Last is prescriptive, where you're like now?
00:24:54.673 --> 00:24:58.505
This is what you need to do to make sure this doesn't keep happening time and again.
00:24:59.420 --> 00:25:11.490
So it's a journey of change management that needs to come through, where we first establish confidence with saying we can tell you always what happened, where it happened, when it happened.
00:25:11.490 --> 00:25:19.026
Once they're comfortable with knowing that and keeping a pulse of everything in the past, they would want to know the why.
00:25:19.026 --> 00:25:23.565
Once they're comfortable with the why, they would want to know when will this happen in the future.
00:25:23.565 --> 00:25:26.861
We show that to them a couple of times, saying you know what?
00:25:26.861 --> 00:25:33.944
If you would have onboarded Arcana six months ago, we could have predicted these three or seven churns, because this model was running this way.
00:25:33.944 --> 00:25:35.287
This is how it predicted.
00:25:35.287 --> 00:25:38.310
It did not see any of this data, which already has happened.
00:25:38.310 --> 00:25:49.626
So we're able to bridge that gap to say, if Orkhana was on your system, you could know when these are churning.
00:25:49.646 --> 00:25:55.094
And then we go from there to saying, now that you're able to build that confidence that we are able to predict, we're able to give you use suggestions.
00:25:55.094 --> 00:25:56.883
We'll give you directional recommendations.
00:25:56.883 --> 00:25:57.405
We would.
