Tom H.C. Anderson of Anderson Analytics/OdinText/NGMR fame is perhaps best known as a tireless advocate and innovator of the use of text analytics for insight generation. I’ve known Tom for a decade now and consider him not only a partner in the quest to drive new thinking in the research space but as a friend as well. That is why even I was surprised when he told me that “Text Analytics Isn’t Enough…” as the message of the rebranding of OdinText, the leading text analytics software he and his team developed, to OdinAnswers. Mr. Text Analytics saying “text analytics isn’t enough”?! Was he ill? Had he been replaced with a Pod Person? I needed to know more about this new Tom Anderson!
So, I reached out and since it had been a while since we caught up, I asked Tom if we could do an interview to explore this new positioning as well as industry trends as a whole. Andy Greenawalt, the CEO of OdinAnswers that Tom brought on last year, joined us for the discussion.
Originally we did the interview via video but had some quality some issues so I want to give a shoutout of thanks to my friends at LivingLens for using their tech to transcribe the audio for me. I have only lightly edited it, hence the more conversational structure and tone of the interview.
In the interview, Tom mentions that they have released a whitepaper focused on their new positioning: “Text Analytics Isn’t Enough… Our new mission provides a blueprint for how to gain clarity at volume and velocity. Find out how to navigate the insights arms race, moving beyond shallow insights to Answers.”
Tom and Andy are at IIeX NA in Austin this week, so if you want to learn more I’m sure they would love to hear from you.
(LM) Hello everybody. It’s Lenny Murphy here with another entrant in our interview series, and today it is my great honor and privilege to talk to one of my oldest friends in the research industry; Tom H.C. Anderson the man, the myth, the legend. And Andy Greenawalt, who came in as the CEO of OdinText. How are you doing guys?
(TA) Great, very well. Thanks for having us.
(LM) Yeah, always good to catch up. So Tom, let’s chat for a second because our careers have counter paralleled; we’ve been doing this for a long time and a lot of folks know you from Next Gen Market Research and Anderson Analytics back in the old days, but then you transitioned into OdinText and launching this technology platform. Out of the blue last year you decided, “I’m gonna bring in a CEO and we’re gonna do this differently”, so give me a sense of what took you there to bring in Andy, and then, Andy, we’ll talk about what you’ve done now that you’ve been under the hood for a while.
(TA) Yeah well, I met Andy and you don’t know what you don’t know until you meet someone like Andy who has scaled at several technology companies. … My expertise has been market research and specifically text analytics, text mining. Well, we would say customer analytics, really. We’re very impressed with Andy and his experience and in terms of thinking about how to make Odintext and OdinAnswers more useful for more people.
(LM) Okay … OdinAnswers, we’ll circle back to that in a second. Cause I think you’ve got some cool stuff to say. So Andy, You met Tom and he wooed you; now what’s happened in the preceding almost year?
(AG) Yes it’s been a year, yeah it’s just a year as my LinkedIn anniversary hit. So as Tom said, I’d scaled a number of SaaS companies early on. And I had invented security in the Cloud back in ’97. So I’ve been, putting stuff in the Cloud for a while. And but my academic background was actually in cognitive linguistics and philosophy. And so always having this sort of background thread about text analytics and about understanding at scale. And so met Tom and through one of Odin’s investors. And was incredibly impressed with what had been done, the customers that they served. But realized like hey, you really cracked the code. But we need to build a platform to head into that opportunity that is the next number of generations up. I found that opportunity to be incredibly exciting. And yeah I’ve worked on a number of things on the text side. But they were in different markets and to actually be working on the tech that I understand, in this realm of understanding people. Cause ultimately what this is all about is understanding this intersectionality about people and letting business’ better understand their customers and the aggregate effects that could have, is a stupidly exciting opportunity to have the opportunity and to work on.
(LM) So you mentioned this idea from OdinText, to OdinAnswers. Now first, does OdinAnswers mean you have to give up an eye? I hope that’s not the currency!
(TA) No, we already did that. No, so OdinAnswers for a lot of reasons – there’s a white paper, that we are just releasing now, that explains our thinking and that of some of our customers that we’ve been working with in terms of OdinAnswers – and that is that text analytics isn’t enough for business. And OdinText has a…
(LM) Whoa, wait, Tom, you just said text analytics is not enough, the guru of text analytics. Alright, you gotta explain that more buddy.
(TA) Well text analytics, the way a lot of people have applied this has been in isolation in a silo. You just can’t do that. Customers want answers. And so why would you be running this powerful thing, in a vacuum. And so the contextual data that people want to understand, Obviously if you don’t know who said something, that’s a lot less useful. And then importantly, what impact does it have on the business, the business metrics? And that’s something that we’ve worked hard to really, distill what an answer is, the best answers for companies, and working with them to provide that scale now.
(AG) And this was one of the interesting observations as an outsider. I kind of came in, I’m looking at the insights and the answers that customers are getting with the platform. And it’s really like we’re almost, we’re almost mislabeling it as text because it’s not just text, it’s this intersectionality. So let’s, let’s stop talking about the meat byproduct reassembly system, that’s kind of the text. The sausage, that’s the answer. So let’s actually get the emphasis on the right, the right sílabo, and focus on answers. Because that’s really, cause when we shifted our, began shifting our focus there, you realize like all of the opportunities for increased contextualization, increased like the technology that can be thrown at getting to managerially ready answers. Because that’s, that’s the valuable part. Just the input is only so good. It leaves, it leaves a, it leaves a long last mile problem.
(LM) So let’s get into the specifics a little bit on that. So if we think about, from my own experience watching your evolution, I would think of almost every solution dealing with text analytics and social media as a utility, right? It was a way to ingest information and to put that out, but there wasn’t much context around that from a business issue alignment perspective. I think that’s that gap that you’re talking about Andy. It’s a great way to take instructive information, make it manageable and produce information. But to get that information into something that was aligned to answering the business question was still a gap. That was primarily done by a human analyst. If I’m hearing you correctly, then it sounds like what you have done, is actually tried to deal with that gap from a technological standpoint; to allow clients to compare the information coming out and align to internal metrics or KPI’s or business issues rather than just a dump of information. Is that an accurate understanding?
(AG) Yeah, it is absolutely. Because one of the things that, every business, certainly every business that we deal with, they’re selling something to someone, whether that’s a product or a service or an experience or whatever that is so, so it was critical for us to think about, okay what is the, what is the business information architecture if you will, that our clients have. Because all of this unstructured data is talking about something. It is the customer experience’s relative to the nature of the business. So my silly example is , if you’re looking at unstructured data from Ford motor company, is Focus a noun or not?
(LM) Right, right.
(AG) So understanding the business’ conceptually, what is their structure and their core, when we surface an insight out of the data. Does that belong to a product manager or a marketing manager or a support channel? And so by having by taking the if you will the aggregate data set within the platform and beginning to create almost a model of the organization, which then this information can flow against, allows you to cover that last mile. Right to get to, to get to managerially ready answers.
(TA) And basically, what we’ve seen is, I think we are all aware of this that there’s a insights arms race going on right now. And so not are we just flooded with data everywhere and not enough time. And management is asking for answers quicker but at the same time, we’re being flooded with analytical solutions also. And we need to make it easier basically for our users and for our that these companies that we work for. And so that’s what’s, what this has been, been about. And something that we go into more on the white paper.
(AG) Yeah, cause one of the observations that we’ve had was that, text analytics had, often was just a feature, it was a feature embedded in something so, hurry, create work clouds, create some sentiment, do some things but, survey platforms would have something and social listening would have something and they would all have peace corps review data. There wasn’t anything for synthesis and understanding across that. Feature embedded in something–utility. Create word clouds, create some sentiment, do some things. But where survey platforms would have something, and social listening would have something, and they would all have piece parts, review data, there wasn’t anything for synthesis and understanding across them. And it’s really if you’re going to provide answers, that synthesis layer is absolutely critical. So, what does that mean? That means a rate of ingestion and a data normalization problem. And so, getting those things relative to the business model that we talked about earlier, those things conjoined are really the … That’s the technical challenge. And really, that’s the pivot of the business, is engaging in that problem to solve, as Tom put it, the insides arms race. Because in the absence of this unification, how does anyone keep up?
(LM) When you look at the hierarchy of the solution stack, where I would say that a year ago OdinText sat alongside a variety of other utilities to collect and do some analysis of data, your survey, maybe some of the behavioral data, etc. and each of those was somewhat siloed in terms of the analyses that they generated. That flowed up into something else. For the sake of argument, let’s say it’s a Watson, or whatever the hell type of analytical—
(TA) Well, just a plow, yeah, maybe.
(AG) More often than not it flowed up into a group of people.
(LM) Well, fair enough, fair enough. Somebody synthesizing–manually–this information.
(AG) There was some meatware problem.
(LM) Right, that’s right. Now, the evolution is that although that utility layer may still exist, because I assume you are still ingesting data and analyzing that, that’s a core piece, but now there’s another layer of the solution stack that makes it closer to replacing the meatware component that is synthesizing the information from multiple sources. So I assume now that that model allows the synthesis of both structured and unstructured data, based upon the business issue, right?
(TA) Yeah, absolutely.
(LM) Okay, all right, that’s very cool.
(TA) It’s something we’ve noticed as we were trying to look at our client, our users, they really were all over the place in many ways: market research firms, clients in almost every industry, including nonprofit. But the one thing that struck me was how many digital, what we call digital-first companies that we’re working with, which would be counterintuitive, because these companies have some of the smartest data scientists in the world working for them, and yet they were coming to us. Having this more holistic viewpoint, I think, has helped that, because they see the value in that. And they’re looking to get the answers quickly, they’re not wrapped up in the detail, because they got that solved.
(AG) Yeah, the digital-first companies, the YouTubes of the world, the eBays of the world, Ubers, they have their customer experience is their own, because it’s digital first. The amount of data that they have is incredible. It’s so much different than the CPG companies, or anything that existed before. And because their business models can respond in near real time, they need these insights. So they both have the data to produce the insights and have the feedback loops to use the data, to leverage the data in completely different ways. That’s where we found the digital-first companies were coming to us. Which is awesome because working with them is incredibly exciting ’cause you’re not dealing with a six or 12-month loop where maybe something will find its way through management. They’re using the insights in the next month.
(LM) And it’s the important contextual insights, right? Yes. I mean, that is the power of text analytics is to get to the variety of the non-conscious, the contextual, the implicit, all of those things that we try and capture in other ways, but humans just leave as our exhaust and that’s important. But those companies really don’t have access to that. They’ve got the comments, but what the hell do they mean? I have certainly been a fan for a long time that that’s vital for the industry to do. I want to be conscious of time. We are doing this right before IIeX, so you guys are about to be in Austin, and you’ll have a chance to talk to a lot of folks there. And the white paper is released, and we’ll have a link here in the interview. Here’s your chance to give a shout out. Why should somebody take the time, out of 1,500 people at IIeX, to come to talk to you?
(AG) We have two different things. One is the brand customers, and the other one is the vendors and other folks. The vendors and other folks, obviously as you can probably tell from this, our strategy processes, not project, and that’s all about integration and connection. Part of the platform is about integrating and connecting, so we want to talk to as many folks as we can about making those connections, getting that survey data in, getting the review data in, getting whatever data we can get in. We’re all smarter that way. On the brand side, again, if this is compelling, it’s all about finding these intersections, finding these answers. Many researchers that we talk to complain about flat data, ’cause they’re using traditional tools that don’t look at these intersections, and therefore they don’t really know why their NPS has or hasn’t changed. They don’t know because it isn’t digging into the micro-segments, it isn’t digging into the metrics in a way to bubble up where the real opportunities or risks lie. If you’re frustrated with work taking too long, and ultimately not getting to interesting results, then please, look for OdinAnswers.
(TA) And we’ll be having a hosting on a round table on the 23rd to discuss answers, really, from the way we’ve been looking at it. And we’re very curious to hear how customers are looking at it because it’s a really interesting thing to ponder. And so that’s what we’ll be discussing at the round tables on the … Is it the 25th, I believe? Yeah.
(LM) Okay, very cool. Guys, congratulations on this journey. Again, for watchers, this feels a little more personal, just because, Tom, it’s almost like we’ve been … We’ve been along at this together for a long time. It’s really exciting to see how you guys have developed and this next opportunity. And for what it’s worth, I agree. When we talked brands and I look at the industry as a whole, and I think I vlogged about this recently, is the idea of a full-stack solution. That’s where we have to get to. For what it’s worth, I approve. This is a Lenny-approved strategy. I think you called it exactly right, and it’s the right way to go, and that’s where the value really lies. So, congratulations.
(AG) Thank you, Lenny. Thank you very much. Thank you for having us. Enjoy Austin, I won’t be there, so go and enjoy it for me. Will do. We’ll miss you.
(LM) Absolutely, all right. Thanks, guys, talk to you soon.