Evan Liang – Improving the Customer Journey through Lead Routing

Evan Liang, CEO of Lean Data talks about getting agile with lead routing and their massive stack including: Builtwith, Datanize, Yesware, Outreach, SalesLoft, Persist IQ, Trello, SFDC, GoToMeeting, WebEx, Zoom, Capterra, UberFlip, DocuSign and Marketo!

Listen to the Podcast

Read the Transcript

John: Hello and welcome to Stack and Flow, I’m John Wall.

Sean: And I’m Sean Zinsmeister.

John: Today our guest is Evan Liang, the CEO of Lean Data. Evan, thanks for stopping by today.

Evan: My pleasure, glad to be here.

John: Sean, the biggest thing of course – Dreamforce is in your backyard and will be taking over everything in sight so are you prepared and ready for everything you’re going to see and check out at this huge, massive event?

Sean: Yeah, on top of the fact that there’s always going to be that last minute stuff that always looms toward the end, right? The printouts that you need to make and the rush to the copy center. Any of the last minute design and messaging and training stuff, so it’s all leading up to the finish line. I wanted to dive more into the Dreamforce, but you actually, a week or so back, were at Jeff Pulver’s conference, MoNage. They were doing some really interesting stuff with AI and Chat-bots and things like that. Tell us what was going on over there?

John: Yeah, that was quite the event. Jeff Pulver, if you’re not familiar with him, he was one of the early guys in Vonage. The company that eventually was sold off and became huge as far as voice on the web. In fact, Vonage is V-O-N, Voice On the Net. He’s moved to the next thing which is messaging, so he’s come up with MoNage which is Messaging On the Net. This was the first event that he ran here in Boston a couple of weeks ago. It was just amazing, I really was blown away by everything that they were rolling out. There were three key ideas:

The first one was, just the prevalence of messaging. It’s interesting that everybody’s looking to China for this with what they’ve been doing with WeChat. It’s amazing, with the entire mobile market, WeChat just drives everything. You can book flights, you can get a ride, order food, transfer money. All this stuff that people are used to using for Apps is now being done on this chat platform.

One of the stats that they rolled out was that in the past year, the average number of new apps downloaded by mobile phone owners in the US is zero. We’ve reached the saturation point where people really don’t go out and get apps unless there’s something specific they want. Messaging was one part of it and then of course, bots, the ability to automate the chat that happens. A lot of that in the China model is just they throw labor at it, but the third portion was AI, where you start to build natural language processing where these Chat-bots can handle a lot of the conversation on their own. Between the three of those it really was quite an amazing look at what the next ten years might show to us and where things might go.

Sean: Yeah, that’s really interesting. I think that you saw, especially following the Ignite conference, that was Microsoft’s kind of big spotlight and AI is obviously a big hot topic for them. Not just adding AI into dynamics, but they’ve been really big on using AI as the assistance and how that’s going to work. I think we’re going to be entering this really interesting conundrum of finding that balance between, where do the humans use AI to solve these problems. I can totally see very simplified natural language processing being at the front lines and being able to tackle some really quick things, but then quickly trying to escalate to a human sort of tackle some of that complicated stuff.

Having friends who have not only been working in the messaging apps arena, but having several colleagues who have also done intelligent voice recognition for a company called Vlingo, that was actually Boston based, sold to Nuance. Just talking to them about just how difficult this stuff is to actually process and get right. How many times, John, have you ever talked to Siri, and gotten either no reply or completely the wrong thing? That’s a case study in itself.

John: Yeah, I wonder if it’s one of those things where it’s always going to be terrible, and until then suddenly one day, it is going to be working. Definitely loop Evan in here, are you watching mobile at all, and what are you thinking about on this front? Anything that’s come to light recently?

Evan: For our stuff, for the AI perspective, we’re not so much focused on mobile, but we are definitely looking at different ways that we think business’s need to optimize. We’ll go into this a little bit later when we talk about our stack and flow, but we definitely feel that the buyer experience and understanding that there’s a lot more optimizations around the workflows of how people are doing.

Today, I just don’t think the B2B guys are thinking that great and don’t have the data really to try to optimize their flows, so they’re trying to copy each other and follow the latest trends, rather than being very intelligent about how they optimize, how they work and how they interact with their clients. We see a lot of opportunities, first, from our perspective understanding the workflow and making the data better, and then layering on AI to make that a much more intelligent flow. That’s the area that we focus on, a little bit less than mobile, but more on the traditional workflow and how people, sales and marketing, interact with their clients.

Sean: Evan, you’ve been now a Dreamforce veteran for a few shows now, a few years back. How is this one shaping up in comparison to some of the past couple of years? We always had those big themes around this show, like it was the Internet of things, the Social Enterprise, and this time it’s AI. What are your thoughts going into Dreamforce, I’d be curious to get your take on it.

Evan: I think that in this point in time, having seen Dreamforce a number of years, it is just similar to the past. I think you had mobile a few years ago, and like you said, Internet of Things, and last year you had Reporting with the Wave Analytics and those type of things. It seems like the flavor of this Dreamforce is really Einstein and around the AI. I think it’s really exciting to see that Salesforce is embracing that model, I’m curious to learn more, because from what I’ve seen, from what they’ve showed, it’s very, very basic. It seemed like the idea was, hey, let’s do AI and just apply it to every single cloud. I think they have some interesting ideas, I just don’t know how much of that is going to be made into reality and be product-ized. I’m in a wait-and-see mode around that announcement.

Sean: I’m pretty much in the same boat there too, and it’s very interesting to be watching with the Microsoft narrative almost reactive to really positioning around AI and how Google has really started to come out to talk about machine learning first. That’s an interesting idea that I’m curious about your take, Evan, I don’t see AI as really a feature, it’s not something that you just sprinkle on an existing product. You almost have to cannibalize what you’ve already done to really make that a pure sort of AI-first offering. This AI-first, versus AI later, do you think that there is a big cultural change that needs to happen or what’s the balance that’s struck there, especially from a product view.

Evan: I think you really have to understand that the AI is only as good two things: the data underneath it, and the people doing the modeling on it, to apply to it. You just can’t turn on an algorithm and suddenly going to have it spit out the right answers. For us, we spend a lot of time looking at the data, and our background is really around understanding data, understanding the gaps in the data, especially in the Salesforce system. We just think that flying a random model on top of bad data and bad processes is just not going to turn into the right results. I applaud folks saying, let’s go AI-first, there’s a lot of work that needs to be understanding and being very articulate about the problem you want AI to solve.

The industry in the early stage of understanding that, and for the folks who’ve been around and try to solve these problems, like the Infers of the world, where they’ve really had to tune their models and really understand their customer needs, I think that’s the level of customization and specialization that you need in order to make it really work and give you the results that are insightful, versus being like, oh yeah, I knew that already.

Sean: That’s the sort of scaling issue that is going to be very apparent for a lot of these companies jumping in the game. The other thing I think you’re spot on about is that it’s really starts with nailing that quality of the model and how accurate that model is. I’ve heard a lot of bizarre similes to, oh yeah, one model versus another is similar to looking at marketing automation where it’s this deliver-ability, versus that. Is one better but yeah, it doesn’t really matter. No, no, it completely matters! Are you kidding me? That last mile stuff is the automation stuff, the model and building that trust and making sure that things things don’t fall down, or heaven forbid, are mislead is huge. Are you in the same camp, Evan?

Evan: Definitely, we spend a lot of time looking at the data and it’s very sensitive, right? Depending on how you weigh things, and how you build that model, you can have vastly different results. This definitely needs to be very customized, very tuned, so that’s where I have a little bit of skepticism around the Einstein stuff at the get-go where it’s apply a single model across all of our customers. I think it’s going to have to be very tuned and calibrated per customer and that’s going to take time. I am excited that the industry is moving that way and that we will get better and better insights as people spend more time doing it.

John: I knew that we’d hit Einstein within the first ten minutes so we can mark that off the Bingo card. As we look at stuff, Evan, let’s take a step back. Tell us what Lean Data does, what you do, and fill us in on where you are coming from.

Evan: Lean Data is a company who really focused on solving the problem around lead routing in B2B enterprises. Why that’s such an important problem is because lead routing is really about tailoring how you interact with your customers. If you think about it, routing is all about matching your customers when they come to you or you’re going outbound to them with the right person in your company. We feel there’s a lot of opportunities to make those interactions and each of those touches a much more meaningful, more personalized, relevant and efficient than it is today. That’s really what Lean Data is about, solving, we are doing it through routing, but we really think the impact is across your entire buyer’s journey and the experience the buyer has with your company.

Sean: Evan, you get a chance to really work hands-on with a lot of the hundreds of businesses that are customers of Lean Data, and one thing that I’m really curious about, you guys are in the rough stuff. In terms of the state of the stacks that the companies are working with today, what’s sort of the grade, like where are they at with sophistication? Is it like, you guys are really solving a big mess or are they doing okay? What would give them as a grade?

Evan: I’d probably give the people, overall, a ‘C’ right now. I think there’s two areas they could really be better about. One is really optimizing and leveraging their vendors, we’ve built up the sales and marketing stuff; there’s a lot of people just buying a lot of shiny, new objects and not really understanding how they fit into their process. I think there’s a lot of, “hey it worked for XYZ company and let me just buy that and try the same process.” The first thing people have to understand is that everyone’s business process and how their customers buy from them, or their industry, or their price points are going to be different. You have to customize everything, configure everything, for what’s going to work best for your good, or market. There’s that first level of sophistication which is better utilization of everything you have in your own stack.

The second area that I think vendors are going to need to understand a lot better, or our customer, is how you integrate and have the different products in your stack work together. How does the one plus one equals three? That’s the part that I’m excited about that we’ve been collaborating with Infer around, is we have joined clients and we are trying to integrate our two products, and often we see when we make that integration really smoothly, we come up with a product that neither of us had beforehand, and it solves client’s needs and delivers so much more value. The stack today is very silo-ed, and it’s the most basic integrations, with thinking about how those touch points can really deliver even greater value. That’s why I would grade everyone around a ‘C’ today, because there is just so much more opportunity, both in how they can customize it and how they can get the integration the best of the multiple vendors in their stack.

John: What are the big mistakes? When you come into a new customer and they’re saying they’re having a problem getting the round robin right, or the data is just going into a black hole. What are the biggest and the ugliest things that you face, and that you get fixed for customers?

Evan: The biggest thing is really understanding that you are going to have bad data, you’re going to have duplicates, you’re going to have mismatches, you’re going to have missing data. For us, the first part of what we solve for people is starting with the data, making those matches our first initial issue to really be good at account matching, which is this flaw in the Salesforce system that leads don’t match into accounts. The first thing we solve is making those relationships passable inside their Salesforce system, and once you get those connections, then let’s figure out and design the optimal process for them. Understand that it’s an iterative process, you need think outside the box a little bit about what is it that your customers really want and then design the process, versus “I saw this work somewhere else, let’s just use that.” It may not work as well for you, so we try to work with customers to understand your problem set, what is your ideal customer experience, and then let’s design a flow. That flow will be different from every single customer perspective.

Sean: You mention data quality, obviously something that we deal with a lot on the Infer side, Evan, the state of CRM’s that you walk into… we’ve seen a great mixture of things. In terms of some of how they became so poor in terms of data quality, how did it happen? How are companies falling victim to this? Is it just bad hygiene practices or is just bad data input, how do you think this is coming to pass, this that has created a lot of these great problems for us to solve?

Evan: Yeah, it’s just a fundamental state of the technology of how Salesforce was architected, back when it first started. If you think about it, Salesforce is fantastic, it has a great work flow, and it’s kind of an industry standard, but really it’s just a database and a lot of what it is required to move data around and make those linkages is manual processes. Requiring mostly sales reps to do a lot of this data, and as all of us have interacted with sales reps, they’re not the most process efficient, detail oriented, rigorous people who want to spend all of their time making sure their data is right. It’s almost an inevitability based on the way that the system was originally configured and that you are relying on manual people to do that. Both our system, and the AI, that’s where the opportunity is the automation and the algorithms to make these smart linkages and not just rely on people doing it. If you do, every single system is invariably going to get into a messy state.

Sean: I think that is something that everybody can be excited about where AI will be applied and where I am hoping that the direction of Einstein goes, tie it back there. The more that we can use AI to remove the mistakes made and the redundant data entry, is going to improve the overall data quality, help improve the models and then that last mile stuff, that automation, that orchestration, becomes a lot more interesting. In terms of where I wonder that it’s going to start, I know that Salesforce is interested in doing things like Lead Stages and automating how those are changed. Are there other places that people should be looking at hygiene first or are there things that you’re seeing could be automated first beforehand.

Evan: Yeah, I think a lot of it is, and that’s where they’ve talked about Einstein is, and Infer started, is really around the scoring, the attribution, anything that is on the reporting side; those are good places to start, or after the fact. That’s where I don’t think the sales reps want to spend much of their time around, you said Sean, if your too fast into the automation, without cleaning up the data first, you could run into trouble. As soon as the sales people don’t trust the process, or what the automation is doing for them, then they’re just going to discount everything that comes after that. It’s a good place to start really, around attribution, scoring, reporting stuff, where you’re summing the data before going into automation.

Sean: Your system, of course, interfaces with Salesforce, and you’ve got other sources feeding you data, we were talking, before we started recording, about major differences between these open and closed systems and how now it’s so important to buy tools that allow you to get data in and out and integrate with other points. Is that a challenge you guys face, as far as trying to integrate with systems that are less flexible as far as getting data in or out, or what’s your approach to working with the whole stack?

Evan: The great thing is that we’ve made ourselves as native as possible, sitting inside Salesforce. We can see everything that’s inside Salesforce, and be very, very flexible around that. Integrations with other vendors is really if you put it inside Salesforce, we can pick it up, we don’t have to actually integrate with yourself. That works well for a lot of our clients, where they really view Salesforce as the master system of records, the Holy Grail of truth. Being able to leverage Salesforce’s platform and what they’ve done there is very, very smart. They built this huge ecosystem, kind of been open and so that has allowed for a lot of easy integrations with the other partners in the stack. That’s really our philosophy and similarly, everything we do, we leave an audit trail and leave the information in Salesforce so anyone else can pick it up very easily.

Sean: What about on the LeanData side, what does your stack look like? What are some of the tools that you are building out for both your sales and marketing teams?

Evan: We are very account based, and where a lot folks are moving towards, I would describe our stack as first starting with how to figure out which are the right accounts to go after. We play with a lot of vendors, we partner with a lot of vendors, so I’ll just name the various folks and different components. A BuiltWith or Datanyze, helping you get those initial set of accounts and putting those into the stack. Going from there, we obviously look at a lot of… we are very much inbound/outbound so we do look at a lot of the sales cadence stuff, whether it be yesware, outreach, SalesLoft, or PersistIQ, we’ve played with multiple systems, and are tuning those systems to get better and better at doing that. We obviously use ourselves for the routing, and the orchestration, we definitely eat our own dog food.

The sales reps also use Trello for managing their deals, both inside and outside. We are using Salesforce and Marketo for looking at marketing automation, but they also use a number of other free tools like Trello. Meeting set ups, I would say, would be like a GoToMeeting or a WebEx Meeting, or Zoom are various technologies that we’ve tried out.

Towards the later stage in the stack, we have had some success using kinds of like content management, and seeing content as a really great way to nurture and bring folks along. Vendors like Capterra, or UberFlip would be examples of you actually use the content more strategically to nurture people. Then at the late stage, obviously, from order to cash perspective, we’re a little bit too small for a Steelbrick or Apptus but definitely I have looked at stuff like Tinder (actually Tinderbox, which is now Octiv – Ed.), we’ve used DocuSign for signatures. I’d say from the top of the funnel towards the close process, that probably be how I’d describe our stack today.

John: That’s a wide array of stuff, that’s definitely interesting. Trello just jumped out at me, is that more because it’s worked so well for collaboration and for sharing more cards? How did that fit into the mix?

Evan: Yes, that came up from our reps perspective. They just like the visualization and being able to organize their opportunities and accounts that way. It being a drag and drop interface, this really worked well for the millennial generation where they’re just used to a much better UI than the list based way Salesforce. They are literally spitting those things out into Trello and moving them around to do their stages. They find that to be a much better representation than looking at a list of accounts, or a report natively inside Salesforce.

Sean: That’s interesting, because when I think of when everybody thinks of Trello, they think about project management. That has to be something I have to ask about, because I always see project management, the ability to get stuff done as such an important part of any organization, big or small. Do you guys have some sort of philosophy or technology to support how you are doing project management? How do you approach that at LeanData?

Evan: Do you mean project management on the sales side or project management on the development side?

Sean: I would actually be interested to see if it’s different in different departments. If I look at we do here at Infer, and what I’ve done in the past, it’s all Jira, all the time, on the product and engineering side, versus we’re Asana on the sales and marketing side. I’m curious if there’s different flavors, depending on the department.

Evan: That’s interesting, it is different for different departments. I would say on the sales and marketing side, as much as possible, even though I cited Trello, we try to codify as much as our process in Salesforce natively as much as possible. Salesforce and then Marketo, we do have a lot of folks that are really good at marketing operations, and so we try to automate even the sales processes there around that, or your cadence technology. So we’re trying to do that natively, and prescribe that into our sales and marketing system. Per your point all around the product management side, yes, they Slack for the communications, they Jira or GitHub to manage the bug lists in the project management. I would say we are probably using too much Excel these days, or Google Docs to do a lot of that stuff. Certainly could get better at those processes, but I think our stacks is very much similar to other people on the development side.

Sean: That segues nicely into the flow part of stack and flow. Even though you’re a fast growing company, so many tools and I’m sure you that you’re seeing this in other stacks, what are some of the best practices? What should people be looking at, and I’m also just curious about, is this the stack of the future? Do you think it’s going to be this orchestration of many different tools of sharing of data, versus this kind of closed system to go back. It seems like a lot of companies are adopting that idea where you’re using Slack for messaging, Trello for this, and all sort of this stuff. Is this the future, is this just going to continue?

Evan: I believe so, I think the reason for that is because integrations are so easy right now that you can get best of breed in each of the different areas. What I cited earlier, is I think everyone’s stack is going be different and how they leverage all the tools in the stack is going to be different, so that you are going to need best of breed in certain areas. Depending on how you are trying to orchestrate your flow and what your buyers really need from you. That’s why you are going to see some folks who are going to need the best of breed in predictive, in routing, in these types of things. They want to be able to easily integrate that with each other, versus, I think that there isn’t going to be a one-size-fits-all for everyone and limiting yourself because of that.

In the old days, where data didn’t move very seamlessly, it was easier to have a single platform because it allowed the integration to be easy at the trade-off that you weren’t getting the best of breed on top of that. In the old days, when you were in the enterprise software world, the SAP and Oracle world, it was just easier to buy their next module than having to integrate with someone else. Today, where integrations are very seamless and very easy, you are going to be looking for best of breed and the reason is what you configure your stack to do is going to be different from the company next to you.

The analogy that I like to use, look at retail, every store, depending on what you’re selling, a grocery store layout, the type of people they staff, where they put things is completely different from Nordstrom. Even in electronics, how Apple has been really successful, is very different from Best Buy and you can’t just take one-size-fits-all and move it to another person. For example, the guy who was really successful setting up Apple’s retail store, when he moved to JC Penney, it was a colossal failure. You really have to tune everything in, and everyone’s stack is going to be different because of that, in my mind.

John: Evan, once the system is in and up and running, what are the more advanced tactics. You’ve cleaned up the data, you’ve got it so there is some flow working, but where do you go from there as far as getting more value out of having your assignment correct? What accelerates it, once you’ve got the basic tasks done?

Evan: The last thing is the tuning it, generally, what we’ve seen is because routing was something that was very hard to change in the past. People usually only changed their flows maybe every six months, every year, at the annual time when they’re doing the next year’s planning, whereas our best customers have realized that because we’ve made it so much easier for them, they can start changing it every quarter, every month. Once you get all those things in, it’s constantly tuning of that system and just being much more agile and adaptive in meeting your needs, and optimizing it for your own experience. That is the next stage, after that, I very much believe once people have that set, that’s where you start layering on AI, where then your humans have done the best they can, let’s layer on intelligence and algorithms to even better tune it and give insights that the humans may not be able do themselves.

Sean: The one thing that I’m always curious, especially from the human side, is there any best practices around tech adoption? Whether you’re bringing your new technology into an organization and how you coach that organization, both from a sales and marketing standpoint, and maybe how is that mirrored on how you guys treat new tech and how you’re building adoption among users for LeanData as well.

Evan: Two things, one thing for us a big practice of what we’re building is subject matter experts, as we have done hundreds of implementations, we’re taking the learning from each implementation, building best practices guides around that. We share those with our customers, but obviously letting them tune it themselves to make sure it works for them. In terms of optimizing your stack, and what we do internally, we’re constantly trying to measure it and seeing if can attribute things to be actually working. There are definitely things in our stack that we haven’t been able to attribute success to it, and therefore, stop doing it. We have to be constantly re-evaluating our stack, looking for what works, what doesn’t, and being willing to rip things out, and not keep it in there because we thought it was cool tech. Back to basic principles, looking at the data, making sure that it is delivering the value that you think it is, it’s really important about constantly adjusting your stack.

Sean: That’s really interesting, especially having that pruning process, when you are evaluating new tech and things like that. I know that you guys have really moved more into this lead management space and born out of that ABM idea. So many of the new ABM tech out there, do you feel that this is something that the buzz wears off and people come back. It finds it’s place in the marketing mix, or do you continue to see ABM tech stand alone category that continues to grow.

Evan: It great that ABM and everyone’s recognizing it, but I think it’s part of an overall lean management. What I was saying earlier around the buyer experience, ABM is a great methodology, but it’s got to be taken in as part of your overall strategy and go to market and how you interact with your customers. That’s where I see it fits. The example would be a few years ago, when all the rage was inbound, ABM is a facet of outbound. From what I see, they’re not totally separate, I often see that outbound generates inbound, or inbound should be a trigger to do outbound. I think Jason Vargas at Datanyze was really good at saying, one of the best things he did around being account base was to recognize that these two flows had to work well together. ABM is a great mantra, but then people have to adopt it for what they’re doing, and overall they need to think about their end experience for all of their customers, whether or not they’re leads, or accounts, or how they come in: inbound versus outbound. That’s where I see the more holistic story, the longer term goal, and we’re all trying to make our buyer’s experience better. I don’t think anyone can argue against that.

John: Evan, that’s great! We appreciate you taking some time to explain all this to us. If people want to learn more about Lean Data and what you do, what’s the best to get in touch?

Evan: Definitely feel free to visit our website, www.leandatainc.com, the other area that we are trying to build best practices, especially at Dreamforce, we have a new forum called ops-stars.com, and that’s where we’re working with a lot of vendors, like Infer, to build up best practices for our folks. If you are at Dreamforce and need a chance to take a break and get a free drink or food, please come and visit us there.

John: Sean, how about for you; besides logging ten thousand steps daily in the first three hours at Dreamforce, how can folks get in touch with you.

Sean: Probably the easiest is to use Twitter; @szinsmeister is a great way to find me, especially if you’re in and about the Moscone Center, and you just want to find me, come chat about either product marketing tech, sales and marketing, anything, that’s the best place to get a hold of me. Of course, you can always go to infer.com. All of our Dreamforce is over infer.com/dreamforce, pretty easy to remember. If you just want to get in touch and see what I’ve been up to, just google Sean Zinsmeister, and you can find me there.

John: That sounds good; I’m John Wall, you can find out more about me at the Marketing Over Coffee Podcast where we are always talking marketing and tech over there too. You can get past episodes of the show at stackandflow.io and don’t be afraid to leave us a review over on iTunes. That will do it for this week, thanks for listening and we’ll see you next time in the stacks.

John Wall

John Wall

John J. Wall speaks, writes and practices at the intersection of marketing, sales, and technology. He is the producer of Marketing Over Coffee, a weekly audio program that discusses marketing and technology with his co-host Christopher S. Penn, and has been featured on iTunes.

Stay Up to Date