GPS for your GTM

889
03:27:31.970 –> 03:27:35.700
Julia Nimchinski: Awesome welcome to the show. Mark stews

890
03:27:36.550 –> 03:27:37.460
Jonathan Milne: Thanks, everybody.

891
03:27:43.610 –> 03:27:45.979
Julia Nimchinski: GPS for your Gtm.

892
03:27:46.540 –> 03:27:48.110
Julia Nimchinski: Mark Kyle

893
03:27:48.110 –> 03:27:50.379
iPhone (377): Hey? I’m here! I was on mute

894
03:27:51.320 –> 03:27:53.590
Julia Nimchinski: Yeah. For some reason we can’t see you

895
03:27:54.720 –> 03:27:57.539
iPhone (377): Alright. Let me see if I can fix that too.

896
03:28:01.040 –> 03:28:01.660
Julia Nimchinski: Awesome.

897
03:28:02.280 –> 03:28:02.990
iPhone (377): Can you see me now

898
03:28:04.400 –> 03:28:05.170
kylebrantley: Oh, yeah.

899
03:28:05.920 –> 03:28:06.980
Julia Nimchinski: Live, TV,

900
03:28:07.775 –> 03:28:13.920
iPhone (377): Yes, yeah, right? Right, live TV. That’s right. Can you? Can everyone see me now?

901
03:28:14.240 –> 03:28:16.410
Julia Nimchinski: Yeah. Welcome. Back.

902
03:28:16.410 –> 03:28:18.350
iPhone (377): Okay. Great to see you.

903
03:28:18.610 –> 03:28:20.069
Julia Nimchinski: Long time, no see happy.

904
03:28:20.070 –> 03:28:20.870
iPhone (377): Yeah.

905
03:28:21.080 –> 03:28:30.740
iPhone (377): I’m doing fine. I’m I’m actually here in Dallas, giving a keynote at the Smu Cox School of business. So

906
03:28:31.470 –> 03:28:39.130
iPhone (377): I’m happy to step outside and have this conversation. It’s cool

907
03:28:39.750 –> 03:28:42.210
Julia Nimchinski: Fascinating mark. Our community

908
03:28:42.730 –> 03:28:55.560
Julia Nimchinski: is no stranger into your work. And let’s just jump right into it. GPS, for your Gtm, I know that you want to set the stage, and then we can just delve into the demo

909
03:28:56.380 –> 03:29:00.029
iPhone (377): Yeah, I’m gonna put you in really good hands with Kyle.

910
03:29:01.450 –> 03:29:14.040
iPhone (377): alright. So the the context here and I’m I’m aware of of all the other presentations that you’ve already had today and their perspectives. And I’m good friends with a lot of them.

911
03:29:16.150 –> 03:29:19.690
iPhone (377): And it’s all really super important.

912
03:29:20.790 –> 03:29:30.039
iPhone (377): assuming that you have an eye in the sky that’s able to understand how everything actually works.

913
03:29:30.350 –> 03:29:34.750
iPhone (377): So like your post, Julia, in which you said, hey? You know.

914
03:29:35.335 –> 03:29:59.570
iPhone (377): Not a lot of marketing theory actually turns out to be the case. Okay? So using kind of an old analogy that everybody’s familiar with right. That is largely because marketers or sales people or business people, or you know, whoever right they are looking at go to market through a very narrow crack in the door.

915
03:30:00.020 –> 03:30:06.510
iPhone (377): Right? So this is not to mix metaphors. But this is sort of like the blind guys and the elephant

916
03:30:07.010 –> 03:30:15.779
iPhone (377): right? And they’re all on different parts of the elephant. And so they all describe the elephant differently, because they can’t all see the whole thing.

917
03:30:16.660 –> 03:30:18.030
iPhone (377): And so

918
03:30:18.460 –> 03:30:46.289
iPhone (377): this is not a new problem. Things have been around mathematically for quite a long time to help people resolve. This kind of all started out ages ago with multivariable regression. It has then moved steadily forward. The most current most modern offering like this is one of the 4 major types of AI. It’s called causal AI,

919
03:30:46.490 –> 03:30:49.339
iPhone (377): and it allows you to build models.

920
03:30:50.040 –> 03:31:11.680
iPhone (377): That represent as much of the ecosystem of causes and effects in your go to market, in your business, in whatever right that you have access to right, I mean, nobody is perfect, no one knows at all. There’s all the known unknowns and the unknown unknowns right.

921
03:31:11.840 –> 03:31:18.619
iPhone (377): But so it’s a probabilistic calculation, but it’s within the probabilistic framework. It’s very, very accurate.

922
03:31:19.060 –> 03:31:23.550
iPhone (377): The trick is, you know, most people

923
03:31:23.670 –> 03:31:36.309
iPhone (377): they really didn’t like math in high school. They still don’t really like math very much. They don’t want to have to learn how to do it all, and what the principles are and so the bottom line is that

924
03:31:36.710 –> 03:31:45.960
iPhone (377): everyone in my in our line of work is doing the same math more or less, and if they’re not doing in the same math.

925
03:31:46.060 –> 03:31:52.369
iPhone (377): you should find that very concerning cause, that that would be a black box. Right? So then.

926
03:31:52.570 –> 03:32:06.460
iPhone (377): the real issue is, how do you operationalize the math to enable better and better and better decisions by people who don’t understand the math and don’t want to understand the math. They just want to freaking, make a better decision

927
03:32:06.690 –> 03:32:08.809
iPhone (377): than they otherwise would have.

928
03:32:09.100 –> 03:32:22.339
iPhone (377): knowing or not knowing that if you do that enough times you you trigger the law of compounding in your business right, which is highly desirable, and probably one of the most powerful laws in the universe.

929
03:32:22.740 –> 03:32:47.109
iPhone (377): So that’s really the essence of it right? The other thing I just add real fast is that making decisions, particularly in a business context, is navigation. You’re constantly adjusting right things happen. They force you to do something different than than what maybe what you intended to do.

930
03:32:47.700 –> 03:32:56.430
iPhone (377): And this is where the it’s not just an analog. You know it’s it’s actually how it actually really works. Mathematically.

931
03:32:56.960 –> 03:33:01.810
iPhone (377): proof is a GPS, right? And so

932
03:33:02.180 –> 03:33:23.530
iPhone (377): in a nutshell, you have all the things that you’re doing, all the things that you control. You have the outcome that you’re trying to make happen. You have everything in the middle, which is the stuff you don’t control. It’s the headwinds and tailwinds. It’s the wave you’re seeking to surf. All that kind of stuff is holds true. And it’s also time. Lag

933
03:33:23.840 –> 03:33:33.359
iPhone (377): right? And so what you need to be able to understand in order to make a better and better and better decision is, hey, how did these things perform historically?

934
03:33:33.920 –> 03:33:42.119
iPhone (377): How are they forecasted to interact into the future? So this is not being predictive. This is actually a full blown forecast.

935
03:33:42.540 –> 03:33:55.920
iPhone (377): And then you need to be able to automate the model. Recalculation at whatever frequency is appropriate for your business, so that you’re always ready to go. You’re always understanding what’s going on.

936
03:33:56.030 –> 03:33:57.600
iPhone (377): are you

937
03:33:58.460 –> 03:34:04.969
iPhone (377): is is, I mean, like the big problem. I, you and I were going back and forth about this right when we announced it

938
03:34:05.090 –> 03:34:14.849
iPhone (377): right now on a, you know, kind of a typical company. 40% of go to market expense is no longer effective

939
03:34:15.190 –> 03:34:20.910
iPhone (377): that has almost doubled from what it was 6 years ago, Pre covid.

940
03:34:21.720 –> 03:34:24.850
iPhone (377): So that is not because

941
03:34:25.020 –> 03:34:39.590
iPhone (377): marketers and salespeople suddenly took a stupid pill. Far from it, right? That is because they’re not accounting for they’re not understanding and accounting for all of these outside factors

942
03:34:40.050 –> 03:34:41.530
iPhone (377): that have

943
03:34:41.790 –> 03:34:53.689
iPhone (377): conspired over time to negate the effectiveness of a lot of what they have always assumed key for keyword, right? Was effective. And it’s not effective anymore.

944
03:34:54.050 –> 03:34:59.570
iPhone (377): So it’s not an impossible problem to solve by a long shot. And that’s what Kyle’s gonna show you

945
03:35:09.720 –> 03:35:10.490
iPhone (377): perfect

946
03:35:14.250 –> 03:35:15.980
kylebrantley: Hi! Can everyone see my screen

947
03:35:16.110 –> 03:35:17.270
Julia Nimchinski: Yes.

948
03:35:17.510 –> 03:35:18.290
kylebrantley: Awesome.

949
03:35:21.410 –> 03:35:30.250
kylebrantley: So the philosophy behind our causal AI solution was to be able to bring an analytic capability to

950
03:35:30.420 –> 03:35:34.849
kylebrantley: analysts and business users in a rapid and effective manner.

951
03:35:35.530 –> 03:35:40.979
kylebrantley: and philosophically we brought our approach around the idea that everyone has

952
03:35:41.300 –> 03:35:44.700
kylebrantley: critical questions in their mind that they want to answer

953
03:35:45.490 –> 03:35:55.940
kylebrantley: right. If you ask people about what their biggest challenges are from a data perspective, they could probably struggle to describe that completely right because of the complexity and the detail.

954
03:35:56.100 –> 03:36:06.630
kylebrantley: But if you would ask key people, leaders in your business, what are the things that keep you up at night? They can usually rapidly rattle off those top 3 or 5 items

955
03:36:07.020 –> 03:36:16.720
kylebrantley: that are top of mind. And so that’s our was our philosophical approach around the causal AI solution was that individuals would bring questions to

956
03:36:17.461 –> 03:36:30.100
kylebrantley: to bear right for seeking answers to. And this isn’t a big data solution, right? Where you’re looking at tens of thousands or hundreds of thousands of observations to look at potential pattern matching.

957
03:36:30.200 –> 03:36:38.170
kylebrantley: What you’re actually doing is seeking through business questions, to understand a selective set of performance data

958
03:36:38.280 –> 03:36:47.320
kylebrantley: that you can then analyze and look at the causal impacts. And so this is really more of what we’ve described as a lean data

959
03:36:48.010 –> 03:36:56.569
kylebrantley: solution, right? Because lots of business questions can really be answered around a relatively discrete set of metrics

960
03:36:56.690 –> 03:36:58.920
kylebrantley: and observations.

961
03:36:59.300 –> 03:37:03.459
kylebrantley: And so that was the genesis behind the formulation behind proof.

962
03:37:03.600 –> 03:37:07.000
kylebrantley: And we built it around 2 primary persona types

963
03:37:07.502 –> 03:37:13.670
kylebrantley: one being an analyst user. So this is someone that would be an inquisitive, analytically minded individual.

964
03:37:14.140 –> 03:37:21.030
kylebrantley: That would be the recipient of the question from the business, and would go through the process

965
03:37:21.370 –> 03:37:28.510
kylebrantley: of acquiring the data, doing exploratory analysis and completing the causal statistical analysis.

966
03:37:28.910 –> 03:37:37.579
kylebrantley: The other persona is the business user the individuals that would be consuming the results and interacting with the results of the solution.

967
03:37:38.360 –> 03:37:59.999
kylebrantley: And so we had to build the solution to to accommodate both of those types. And, as you’ll see from the navigation side, very briefly, and I wanna highlight these dashboards is our visualization feature visualization layer, where I will be starting. our journey with our demo with and more detail.

968
03:38:00.110 –> 03:38:06.059
kylebrantley: And this is usually the resin area for the business users, right? That they can go in and interact

969
03:38:06.200 –> 03:38:14.800
kylebrantley: through the visualization outputs of the causal analysis workspaces and libraries and exchanges

970
03:38:15.000 –> 03:38:18.830
kylebrantley: are usually predominantly the domain of the analyst user.

971
03:38:19.030 –> 03:38:39.930
kylebrantley: So workspaces are a feature capability where an analyst can sort of wall off an area of specific analysis that they want to do. You would usually organize a workspace around a specific business question and therefore associate the data with that specific business question in that Workspace and do your analytical work.

972
03:38:40.260 –> 03:39:06.700
kylebrantley: Our Libraries house are the persistent assets that we maintain inside the solution. So these are the the data, the correlation, and other statistical analysis assets that are created as part of the of the process that reside within libraries, and finally exchanges is our ability to connect between different licenses of proof. So we can allow

973
03:39:06.910 –> 03:39:15.059
kylebrantley: a brand to securely interact with their agencies through proof and through the exchange feature to

974
03:39:15.310 –> 03:39:42.949
kylebrantley: share data and insights in a secure manner, because we know that this is obviously a challenge with a lot of marketing organizations and a lot of agencies that they struggle to prove their impact on the things that the business cares about, that the business doesn’t allow them to see that sensitive data to, to, for them to be able to show that they’re impacting. You know, sales, for example, and so exchanges, facilitates that in a secure way, all right if you will.

975
03:39:43.370 –> 03:39:46.849
kylebrantley: And so moving into the visualization layer dashboards.

976
03:39:47.200 –> 03:40:16.430
kylebrantley: I wanted to share with you what is sort of the keystone or capstone. This is at the end of the journey of the analysis. So you know, the analyst has already been posed with a question that analyst has conducted their discovery and exploration analysis, and it ultimately results in the creation of what we call a key insights dashboard, right? So this interactive visualization is anchored and based on the notion of a business question.

977
03:40:16.580 –> 03:40:21.190
kylebrantley: And so if I were to look at this simple business to business example.

978
03:40:21.722 –> 03:40:26.699
kylebrantley: and we’re curious about what go to market channel activities are impacting sales.

979
03:40:27.540 –> 03:40:47.899
kylebrantley: So as part of that process. As I said, the analyst will have identified a series of metrics that they had in mind to help them answer this question. So it’d be a series of channels potentially a couple of external factors. And the outcome right? Or the object that they’re interested in, which in this case in this example is sales

980
03:40:48.966 –> 03:41:12.990
kylebrantley: through that process you’re going to see some channels that aren’t going to make it as part of the analysis, and by not making it, I’m saying, not statistically relevant, you know, at this point in time, based on the analysis that’s done. So, you are going to see a winnowing effect right? That the analyst is going to go through all right, ultimately resulting in the model that they come up with to say, Hey, this is the model that’s answering this business question for us.

981
03:41:13.570 –> 03:41:22.080
kylebrantley: Now, when we associate a model that’s already been created into this visualization shell, it gets automatically pre-populated.

982
03:41:22.660 –> 03:41:23.869
kylebrantley: And so

983
03:41:24.220 –> 03:41:29.809
kylebrantley: when we see the causal model that was built B, 2 B Company one, and the factors that make it up.

984
03:41:30.040 –> 03:41:52.139
kylebrantley: And the effect, when this association is done, the 2 primary drivers or factors are highlighted here, so you can see that awareness and sales qualified leads are the 2 primary metric areas or channel activity areas that are impacting sales at this particular point in time, which in this example was at the December 2023, at the end of the year.

985
03:41:52.520 –> 03:42:14.859
kylebrantley: So if individuals are presenting this dashboard, or if this is being consumed asynchronously online by others, they can see their question of interest click on it. And they’re immediately presented with the 2 primary channels that are impacting sales and also any additional contextual commentary that’s been provided by the analyst.

986
03:42:15.110 –> 03:42:23.189
kylebrantley: right? So they can better understand it. So our objective here is showing them what are the primary elements that are impacting sales at that point in time.

987
03:42:23.460 –> 03:42:44.290
kylebrantley: and when you really look at models in the traditional sense, modeling has has consumed the realm of very complex models, right that people are seeking a high degree of accuracy, and there are enormously expensive, but they’re also not responsive. They’re not agile.

988
03:42:44.510 –> 03:42:56.850
kylebrantley: And so our approach with proof was creating smaller models on a much more rapid basis to provide you just in time information so that you can make business decisions more rapidly

989
03:42:56.940 –> 03:43:22.449
kylebrantley: and in most models you’re looking at probably 2 or 3 channels at any particular point in time that are most important or dominating at that time. Right? You’re not going to be contemplating 15 or 20 channels at 1 point, and say, how many levers do I push or switch, or put more in here there? Because you, as a human being, you can’t deal with that. You can’t consume that. So that was an intentional, also.

990
03:43:22.650 –> 03:43:37.920
kylebrantley: Human factors, design consideration that we brought into the solution that we were going to present things mostly in threes or in fives, so that people are able to consume it rapidly and move forward, and in a fit for purpose approach

991
03:43:38.440 –> 03:43:45.609
kylebrantley: now embedded within the key insights. Dashboard is our forecasting feature that we call the route to value.

992
03:43:46.430 –> 03:44:02.380
kylebrantley: So as this dashboard is created during the creation process. All of those factors that were part of the causal model that I showed you in the upper corner earlier are now run through a forecasting algorithm to create a separate forecasting model.

993
03:44:02.540 –> 03:44:13.470
kylebrantley: and then it maps the outcome factor there, which in this case is total sales, because that was the object of the question right that we had that had been posed.

994
03:44:13.780 –> 03:44:25.910
kylebrantley: And so in this visualization, it will generate the historical, actual 12 periods of recorded sales, data that are housed within proof, and then it will generate

995
03:44:26.020 –> 03:44:30.420
kylebrantley: a baseline most likely forecast that’s shown in the red.

996
03:44:31.320 –> 03:44:40.739
kylebrantley: So this most likely baseline forecast has been generated, based on the trend and seasonality factors that are detected in the historical data.

997
03:44:41.020 –> 03:44:52.729
kylebrantley: So you could almost make the supposition that your most likely baseline forecast is if you keep doing the things that you’re doing, you may most likely expect this kind of forecast going forward.

998
03:44:53.360 –> 03:44:57.290
kylebrantley: However, knowing that everything is dynamically changing.

999
03:44:57.420 –> 03:45:04.719
kylebrantley: we wanted to provide a capability for a business user to be able to dynamically interact with this feature.

1000
03:45:05.070 –> 03:45:12.549
kylebrantley: And so we wanted the ability to create up to 3 alternative scenarios here that you see that are labeled s. 1, through s. 3.

1001
03:45:13.470 –> 03:45:28.182
kylebrantley: And this allows the capabilities for the business user to do. What if scenarios in this context. So you can see with the forecast model. In the baseline scenario there are 3 factors that are listed here.

1002
03:45:28.730 –> 03:45:32.479
kylebrantley: and their values as of December 2023.

1003
03:45:32.880 –> 03:45:43.450
kylebrantley: And then, when I go to consider an alternative scenario and I click on that tab, the Ui changes. And so we’ve built what is a Slider bar

1004
03:45:43.670 –> 03:45:50.120
kylebrantley: essentially. And this allows the business users to introduce a growth factor.

1005
03:45:50.240 –> 03:45:54.300
kylebrantley: It’s either increasing positive or increasing negative.

1006
03:45:54.310 –> 03:46:17.099
kylebrantley: So it’s almost like levers like, if I do a little bit more of this or a little bit less of this, and then click, refresh, graph, and then it will generate an alternative forecast curve for us. So in scenario, one in the blue. If I’m able to improve my awareness by 1% per period, which in this use case is per month. Because we’re using monthly

1007
03:46:17.100 –> 03:46:26.860
kylebrantley: time series data. In this example, I could expect a forecast curve that’s shown here in the blue from a sales perspective, a potential most likely salesforce cast curve.

1008
03:46:27.750 –> 03:46:45.129
kylebrantley: And so this arms business users in a language that they are familiar with, which is 1st understanding which channels are effective. So what’s working? It answers that question that management always has for the team when they come in the in the conference room

1009
03:46:45.902 –> 03:46:53.099
kylebrantley: and then the next question is, okay. Now that we know the answer to that, should we be doing more of what’s effective

1010
03:46:53.360 –> 03:46:56.680
kylebrantley: right now or less, or some combination.

1011
03:46:56.890 –> 03:47:12.630
kylebrantley: And so this provides this ability through the use of the Slider bars for people to be more intuitively comfortable with using percentage increases in in driving potentially different scenarios and having those presented on the screen.

1012
03:47:13.170 –> 03:47:24.670
kylebrantley: Now you can drill into these individual forecast, curve, forecast curves, and I’ll isolate scenario one as an example.

1013
03:47:24.860 –> 03:47:51.319
kylebrantley: so you’ll see the most likely forecast curve that’s highlighted in the blue. But you’ll also see a dotted line both on the top and bottom that follow us. And this is also mapping and taking into consideration the error rate in the forecasting model. So there’s a potential range of values that may be realized within a particular month in the future, right by this forecast, and as you look over time you’ll see that the forecast error rate.

1014
03:47:51.350 –> 03:47:59.520
kylebrantley: so it slightly increases over time, which is what you would expect as you’re looking farther into the future. This is a mathematical reality of

1015
03:47:59.610 –> 03:48:02.440
kylebrantley: statistical analysis, so

1016
03:48:02.530 –> 03:48:21.619
kylebrantley: this gives the opportunity for individual business users to isolate individual curves right for further study, and then to see what potential scenario right looks attractive to them, which one do I want to make a potential investment in? Do I want to reduce investment in a particular channel?

1017
03:48:21.630 –> 03:48:44.959
kylebrantley: Right? Say, our direct marketing. Take a small percentage out and move some more into awareness. Top of funnel, because we’re seeing good results with that that’s helping us in the mid in the end stages of our funnel right? So this really facilitates this business conversation. I think that marketing really needs right and would make them far more effective

1018
03:48:44.960 –> 03:49:03.290
kylebrantley: in the executive c-suite, particularly with Cfo, all right, and the financial planning team and their ability to bring this kind of capability to bear and be able to have a causal discussion, right of what’s working, what isn’t and what potential changes they may be able to be making. Going forward.

1019
03:49:03.670 –> 03:49:08.420
kylebrantley: Now another very powerful aspect of the solution. And Mark alluded to. This

1020
03:49:09.000 –> 03:49:14.020
kylebrantley: in his intro is the ability to do automatic recalculation.

1021
03:49:14.560 –> 03:49:34.879
kylebrantley: So when new data is introduced into the proof platform, so say in this example. I’ve now completed my month of January, and we’ve gotten those January data observations available now. So I can update my sales number. What was the awareness for January click through rate SQL, etc.

1022
03:49:35.010 –> 03:49:37.340
kylebrantley: When those updates are made.

1023
03:49:37.470 –> 03:49:48.080
kylebrantley: all of the statistical assets that have been created inside the tool that use that data get automatically recalculated and updated.

1024
03:49:48.240 –> 03:49:56.770
kylebrantley: And so what actually happens is over time. This forecast curve is getting dynamically regenerated every month.

1025
03:49:57.130 –> 03:49:59.720
kylebrantley: right? Again and again and again.

1026
03:49:59.870 –> 03:50:02.470
kylebrantley: And so this is this important update

1027
03:50:03.162 –> 03:50:18.849
kylebrantley: or dynamic element that we brought to bear in operationalizing this aspect for causal AI, where, if you look at more. You know, traditional approaches in the past. It was very complicated models that were very heavy, and we’re not

1028
03:50:19.000 –> 03:50:32.859
kylebrantley: subject or or did not facilitate rapid update or rapid refreshing. And so our purpose was to be available to give people fit for purpose decision making as quickly as possible.

1029
03:50:33.010 –> 03:51:01.599
kylebrantley: right? Because they’re gonna make decisions, whether they, you know at some point have the data or not, right? Because they’re compelled to make those decisions right? The Cfos, the Ceos will make those difficult decisions, even with small amounts of information. Right? Because in a lot of cases, that’s what they do have right. And this kind of approach can bring an incremental improvement in that decision, making process that can have significant, positive impacts on the business and the business outcome.

1030
03:51:02.360 –> 03:51:19.909
kylebrantley: And so this is the feature and key feature that we put in place around our approach of simplicity, the ability to create rapid focus models very quickly, and the ability to bring automated updating to bear right so that you can make those pivots or bets

1031
03:51:19.910 –> 03:51:32.369
kylebrantley: within a fiscal quarter or a fiscal year. And it gives you that opportunity right to adjust and potentially take advantage of an opportunity or to react to a changing condition in the marketplace.

1032
03:51:33.430 –> 03:51:44.169
kylebrantley: And I know I think we’re approaching 25 min in, and I want to be sensitive. If there’s a question section. And so, Julia, I’ll pause here and see if there’s any questions

1033
03:51:44.340 –> 03:51:48.160
Julia Nimchinski: Great Kyle. Yes, we have a lot of questions and

1034
03:51:48.710 –> 03:51:54.250
Julia Nimchinski: every pattern here people are asking, do you have competitors?

1035
03:51:54.860 –> 03:52:00.730
Julia Nimchinski: And they’re they are comparing you with tableau and salesforce analytics. But yeah.

1036
03:52:01.070 –> 03:52:07.619
iPhone (377): Yeah, so so tab both tableau and most of the stuff that is out there

1037
03:52:08.020 –> 03:52:33.819
iPhone (377): that is not actually a competitor of ours, but is seen by many to be. One is based on correlation, not causal inference, math. And there’s all the difference in the world. So to give you a sense of it. If you look at I mean, you know, I’m sure I’m gonna call on this. But you know the the the Google, mmm, the adobe. Mmm

1038
03:52:34.616 –> 03:52:40.220
iPhone (377): sales Meta Robin. Right? All that kind of stuff tableau.

1039
03:52:40.510 –> 03:52:45.940
iPhone (377): These are. These are correlation engines.

1040
03:52:46.270 –> 03:53:03.829
iPhone (377): And in the case of tableau, primarily Bi, right? And so when you’re the difference between a correlation outcome and a causation outcome is usually around 5,000 basis points. So that’s about 50%

1041
03:53:04.340 –> 03:53:05.210
iPhone (377): off.

1042
03:53:05.810 –> 03:53:06.870
iPhone (377): So

1043
03:53:07.410 –> 03:53:19.750
iPhone (377): correlation is really super important. We use it ourselves in the tool as part of the discovery process. But if you end there, you’re in a bad place

1044
03:53:20.880 –> 03:53:21.580
kylebrantley: Yes.

1045
03:53:21.800 –> 03:53:27.490
iPhone (377): And there’s and that again, this is not something that any data scientist would dispute.

1046
03:53:28.160 –> 03:53:31.399
iPhone (377): This is mainstream. That’s a mainstream statement.

1047
03:53:32.430 –> 03:53:41.830
kylebrantley: And in addition to that, as a quick aside, we’re also a salesforce partner, because we have a separate solution that we’re in partnership with salesforce, and we’ve shared this tool

1048
03:53:42.090 –> 03:53:45.030
kylebrantley: a couple years back with their Datarama team.

1049
03:53:45.200 –> 03:54:13.980
kylebrantley: and they acquired datarama. And they’ve been folding that in and said, Wow! This is really great stuff. Guys, you really complement the data Rama analysis because the marketing speak for data. Rama was talking about sort of Ca, inferred causal analytics, you know, in the language, and it’s very clear that they do not do that. And their engineers, you know, seen this solution and said, Oh, yeah, you guys are absolutely complementary to our world. We don’t do what you do

1050
03:54:14.690 –> 03:54:17.579
iPhone (377): Salesforce marketing. Cloud leadership has said the same thing.

1051
03:54:17.700 –> 03:54:28.809
iPhone (377): So it’s not. It’s we we are. We do have competitors. We have legit competitors. But but this, the the ones that we’re talking about right now are not

1052
03:54:31.000 –> 03:54:37.879
Julia Nimchinski: Another question here. How does it adjust to things like seasonality or market shifts, so that the insights stay relevant

1053
03:54:39.440 –> 03:54:40.140
iPhone (377): But you can

1054
03:54:40.140 –> 03:54:40.660
kylebrantley: See it.

1055
03:54:40.660 –> 03:54:41.340
iPhone (377): Yeah. Go ahead.

1056
03:54:41.340 –> 03:54:42.029
kylebrantley: Go ahead, mark.

1057
03:54:42.210 –> 03:54:42.650
iPhone (377): No go ahead!

1058
03:54:42.650 –> 03:54:43.375
kylebrantley: Well,

1059
03:54:44.430 –> 03:54:57.849
kylebrantley: with the, with the model that’s been created. And and the model is the basis for this that it operates off of right. It’s going to be confined around those constraints of the model definition. So as you’re updating this and you’re seeing factors change.

1060
03:54:57.980 –> 03:55:11.439
kylebrantley: the idea is that this is evoking questions right in the discovery process, because ultimately, what we’re trying to create here is an economy of learning experience for people that they’re able to learn in an efficient and effective way

1061
03:55:11.550 –> 03:55:39.230
kylebrantley: about what’s going on. And as you see, factors changing right, it may interpose a question for you to say, Okay, are there additional questions that are now posed here? And do we need to go back to our model and look at evolving the model? Right? Is there a second model that we want to put in place right, that run parallel to this one. Or now, do we know we need to move to a new model with 2 new additional factors that we didn’t consider before right and move forward with that. And so

1062
03:55:39.560 –> 03:56:04.100
kylebrantley: we’ve created this around the idea that you’re going to be learning and changing and evolving right your position on models. The idea is not turning on proof, and that you’re going to have the answer to world peace, or there’s the pressure to have answers to world peace immediately at your fingertips, like I’ve got to get the model, and that is the model for all time. No, this is really in our at the end. For us a learning platform

1063
03:56:04.500 –> 03:56:23.160
iPhone (377): Yeah, it’s this is super important, particularly these days. You know, there’s so much volatility in the marketplace from so many directions that the idea that you can be right today and continue to be right 3 months from now

1064
03:56:23.400 –> 03:56:25.249
iPhone (377): is kind of ludicrous.

1065
03:56:25.470 –> 03:56:33.200
iPhone (377): right? And so so it’s sort of like, you know you’re you’re in your car. You’re driving to a restaurant.

1066
03:56:33.370 –> 03:56:46.130
iPhone (377): The GPS has said, you know this is the best route. This is the way to go right. But then, halfway there, all of a sudden, there’s a big accident, and it’s the traffic piles up, and it’s clearly no longer

1067
03:56:46.360 –> 03:56:51.819
iPhone (377): the best route. And so what does the GPS do? It gives you alternatives.

1068
03:56:52.260 –> 03:56:55.889
iPhone (377): and that’s exactly the way proof operates

1069
03:56:58.590 –> 03:56:59.560
Julia Nimchinski: This is great.

1070
03:56:59.720 –> 03:57:02.860
Julia Nimchinski: We are transitioning to our next session.

1071
03:57:03.040 –> 03:57:13.700
Julia Nimchinski: Kyle Mark, just before we load this up. What’s the best way for our community to support you for all of those watching, and curious to learn more.

1072
03:57:14.251 –> 03:57:16.210
Julia Nimchinski: Where should they go? Linkedin

1073
03:57:16.480 –> 03:57:27.441
iPhone (377): Yeah, Linkedin, ask questions privately, publicly, right? It’s either one is good. And we’ll answer them.

1074
03:57:28.050 –> 03:57:38.550
kylebrantley: And also our website, mark oversaw the development of a nice Llm that we have now as part of proofs website. So go there and ask your questions.

1075
03:57:39.230 –> 03:57:40.630
iPhone (377): Right, absolutely.

1076
03:57:41.110 –> 03:57:49.990
iPhone (377): In fact, it’s it’s 1 of the 1st Llm driven websites. So it’s kind of a it’s an interesting artifact just in and of itself

1077
03:57:50.900 –> 03:57:51.860
Julia Nimchinski: Super cool.

1078
03:57:52.550 –> 03:57:57.240
Julia Nimchinski: Thanks again, and welcome to the show. Cindy Daio.