Attribution Errors

Where did the new client come from? Who gets credit? Should the new business person be involved in growing existing accounts? How should they, and/or the account people be compensated? Blair addresses all of these sales attribution questions.

Transcript

David C. Baker: Blair, today we're going to talk-- [laughs]

Blair Enns: Yes, I know why you're laughing. Again? [laughs]

David: About attribution errors. Before we talk about attribution errors, could we talk about SD card errors.

Blair: Recording errors?

David: Vaguely I remember recording this just last week and then 20 minutes into it, I ask you another yet one more brilliant question and you say, "Oh, I'm not going to answer that, David, I just realized my card stopped five minutes in."

Blair: The listener is not interested in this, David.

[laughter]

David: We probably don't have time to go in when we both went to London and it was a live event and I forgot to press the record button so let's not talk about that, right?

Blair: Yes. That was the first time with that tech, I've used this setup 100 times or 103.

David: Right. Attribution errors. We're going to talk about mainly sales attribution like where does your client come from? When I was thinking about interviewing you on this topic, I just couldn't help but think of all the times I've asked a new prospect, how did we connect? I'll know the answer to that and they'll give me something that I just know isn't true. I also remember back to the days when I used to use Salesforce before I got converted. Having to make a choice around attribution, it's like I have no idea. That's what we're talking about, right?

Blair: Yes. It's funny, isn't it? When somebody buys pricing creativity from us, part of the follow-up the automated email that goes out is, "Hey, by the way, how did you hear about us?" This just came up in a team meeting last week. What do we do with these? Well, I see every single one of them, they get forwarded to me. I don't do anything with it because I think it would be a mistake as the listener will learn as we get into this. I think in a lot of types of businesses, you just need to let go of the need for precise attribution.

It's all helpful, it's all informative, but one of the more left-brained people on the team was saying, "Well, should we be doing some statistical analysis of this?" Maybe we should and maybe somebody could tease out the answer to this question, but most people don't reply because I think they don't know. A lot of people just cite multiple different sources. You and I trade notes from time to time when somebody says, "Oh, I heard about you from your podcast or the thing you did with David Baker these years ago."

We pass those back and forth just to keep each other informed on the topic of sales attribution. You get some weird ones. I remember you sharing one a little while ago that said, "I saw you speak seven years ago, or the new business summit." When's the last time we did a new business summit? 10 years ago, 8 years ago, something like that?

David: It was at the new business summit in 2017. It's like, "Oh, we didn't do them back then." Then even in your own business too, you see this because you're trying to replicate some of the success you've had and you realize, in spite of your best marketing efforts, that the path to get in your door as a client is really confusing. You can't just copy and paste it, it's just pretty messy. You're just accepting that theme and saying, "Okay, what does that mean in an organization especially an organization where some people want credit for something and we have an attribution error." I love this topic. We've touched on it briefly, but not a whole episode on it so I think this is going to be great.

Blair: Yes, and I think the episode was on compensating new business development people so this came up. We probably spent five minutes on it now we'll spend 25 minutes on it.

David: What is the fundamental attribution error?

Blair: When you talk about attribution errors, you start with the big one. I've brought up the fundamental attribution error probably in two or three podcasts. It's a term used in psychology where we tend to ascribe to people's behavior a motivation that is rooted in their character rather than in their context or their incentives. We assume good people do good things, bad people do bad things because it's in their character when the reality is people tend to behave in a way that the context and specifically the incentives incentivize them to behave.

It's not like character isn't important, but the big fundamental error we make. Just take politics. The leader of your team instigates a policy and you look at it and go, "Yes, that's a good person, that's a good policy." Meanwhile, two or three leaders ago, the leader of the other team implemented the same policy and you thought that's bad policy because you are a bad person. That's fairly common.

That's the fundamental attribution error. I wanted to start there. There aren't many days when I don't think about fundamental attribution errors, it's that Baader Meinhof phenomenon again, where once you're driving a Honda Civic, you see them everywhere. Once you start thinking about fundamental attribution error, you'll see it everywhere. You'll see people making judgments about other people based on what they deem to be their character.

David: Would something like that even exist in Twitter? That would surprise me.

[laughter]

Blair: It's possible.

David: I just need to go look again, now that I have this insight. Then the profit attribution error is a little bit different, and it's something different even from the sales attribution error.

Blair: Yes, I've never heard anybody call this the profit attribution error, but I think I was listening to one of Ben Thompson's podcasts, either stir tech re or Dithering, and they were talking about this. What they were talking about was like a reminder of when Steve Jobs came back to Apple, so remember, not too long ago, Tim Cook, CEO of Apple, testified before Congress, I think it's any trust or monopoly, something around the App Store, and he was asked by a member of Congress, "How much profit is the App Store responsible for?" He said, "I don't know". He was pilloried for that response.

That is such an invasive answer the CEO of a corporation, a line of business that generates billions of dollars in revenue, and he doesn't know how much profit? Well, it turns out, that's actually a correct answer because when Steve Jobs came back to Apple in, I forget what the year was, somewhere around 2000. Nick was earlier, the mess that he inherited, according to him, Apple as a company was not profitable, but the heads of every division or business unit claimed that their unit was profit. He made the decision, now we're going to measure profit one way, and it's just one profit. We'll measure impact on revenue, et cetera.

When it comes to profit, it's only the profit of the company that counts. That was such a profound moment for me because I was in the midst of breaking up my own business into two different distinct units where we're moving that training business off to one side and the book business off to the other side, and Shannon runs a training business, and Colette runs the book business. Now I'm just a podcaster. I was thinking about, I know how to attribute revenue, but profits really difficult because it's difficult to attribute expenses, and one drives the other. Then I heard this discussion around Steve Jobs and his decision not to assign profit to business units.

I thought that is such a genius move. That's, effectively what I did is I stopped trying to attribute all revenue and all expenses to both sides of the business. They're related. They're interrelated. At the end of the day, what matters for-profit point of view is just the one number from the combined entity. Even though staffing is starting to line up along those two different entities, and we count revenue separately from those two different entities, we're not attributing profit to any of those two entities, it's really just the combined profit that matters. That was a revelation to me.

David: We're talking specifically about sales attribution.

Blair: We're about to, yes.

David: We haven't yet, right. Stick with us people. The whole attribution thing is pretty interesting. I'll just insert a little story, something that I've noticed frequently in environments where there's not enough structure, I think of it like a little sandbox on the beach and I get credit for everything I build in here, but it's a part of a bigger story, and we're all working together, but at least I work in my space, where you don't have that, in an environment, then people tend to be selfish with attribution, they need to pull things in close to their chest and say, "Hey, I'm responsible for this, recognize me."

People are way more concerned with attribution when there's less structure. The same thing would be true in a sales environment, but here it's monetized and how much money you make can be impacted by whether you attribute the sales to the right person, which it isn't that clean, areas outside of sales, but it's very true, it's a very human thing to want to do this for good and for bad reasons. You're just saying, we need to relax a little bit about it.

Blair: Yes, the two reasons to do this are it's going to affect how we pay people, and I forget the second reason.

[laughter]

David: Blair's been up since 2:30, or something.

Blair: 1:00 AM.

David: Regale his plans on the other side of the Atlantic.

Blair: I said, "No, Im mentally sharp to do this, but I have a short-term memory problem. Let's see if we can get to the end of this second take of this podcast.

David: Okay, so let's just dive right in then and talk about you have multiple elements of this, why it's an issue, how we fix it, what's important, what isn't, just start with this whole data thing. You illustrated this already. You said we have all this data, should we do something with it? It's like, oh, data just points in a direction it doesn't answer all the questions.

Blair: As marketing has become more digital as the Internet has become pervasive, et cetera. You and I have seen this. We've seen this in our clients or listeners that their businesses there was a day back in a creative firm would throw up their hands and say, "Well, I don't know. The ad was great. We're not responsible for the poor sales. I'm being a little bit glib there, but there are lots of ways to avoid the attribution decision to save your position as an individual or an outside agency because there's so may variables and it's so hard to measure.

As marketing moves online, we have access to all this data, the attribution tools and models get more and more sophisticated. You would think the correlation is that with all this increased access to data that we would have solved the sales attribution problem. Maybe there are certain businesses out there that claim to and really have solved the attribution problem. I just don't see it in an agency in any customized services business. I think we need to let go of the pursuit of perfect or even precise attribution and we need to be comfortable-- This is me jumping to the bottom line here. We need to be comfortable embracing the messiness of it all.

David: Now this might work in an environment where you're just doing one thing and you turn it up or turn it down, you see the impact. Most all of our efforts, especially our sales efforts, are coming from many different angles and it's hard to figure out which one of those is most important. Bigger firms that are at scale like SaaS, why are they doing this, though? They're trying a big firm like that would be listening and saying, "Wait a second. That sounds like a sure way to waste a lot of money. Why is my situation different than maybe a salesperson's attribution error?"

Blair: I think a scaled product or product has services business that has thousands of customers in different regions and sophisticated online marketing, I think not only you can get to some sense of attribution. You can make employee compensation decisions based on that because I think the volume of the deal flow covers up any errors. There's an error rate of X, whatever it is, but I think the sheer volume of opportunity is there, and just the scalability of how you go to market, et cetera. It's more scientific.

I think there's just more room in those types of businesses. When you try to bring that model into your own firm, you realize that one error, you get the attribution of one client wrong, and you end up paying one person and serve another person or just not paying a new business person. That can be really problematic. That can lead to bad blood. Ultimately, where the rubber hits the road here, is on this idea of compensation.

This is where the sales attribution issue becomes a problem because anytime you find yourself or you find your team in a situation where somebody is trying to exclude somebody else, and often that somebody is you, you the owner is trying to exclude a Biz Dev person from a deal, to keep them out of a deal, so you don't have to pay them, or you find somebody trying to insert themselves in a deal, like a Biz Dev person trying to assert themselves so they get credit, that's a sign that you're incentive structures are wrong. You're incentivizing the wrong behavior.

One of my most successful friends who runs a firm of well over 1,000 people, I remember talking to him two or three years ago about this. He said, "My salespeople don't make any commissions. All their incentives are based on team performance." He said, "We are a team-selling organization and I want the ability for people to pull in whoever they want to get a deal done." Many years ago, our companies were both companies of one person. He's is now 1,300 and mine's eight. He's doing something right.

David: Other than that, you've had very similar paths.

[laughter]

It's also when you decide to put some attribution system in place, not just to pay the right person, but also to figure out what to do more or less of from a marketing standpoint, what element do you pick? Is it the first time they heard about you? Was it the last thing that they heard that pushed them over the edge? You and I both saw the tweet from Carl just a couple of days ago that said, "Agency owner about revenue attribution. It's not the last beer that got you drank, it's all the beers." [chuckles] I just loved that.

Blair: Yes, that's what Carl's friend of the pod. That was just so timely. That's a last-touch attribution model. There are first-touch attribution models. There's multi-touch attribu-- There are these different models for how you assign the importance of the various elements of your marketing plan. That was the second part that I forgot. That one was lost in my short-term memory.

Sales attribution helps you decide where to spend your marketing dollars. I really wish, and in some businesses maybe they have figured this out, and maybe it's like somebody's SCO or PPC businesses are probably far better at it than the average creative firm. Maybe we'll hear from them on Twitter and there's some learning that we can share with other listeners who were-- At least just convert my point of view on this.

Blair: How you spend on marketing. That's the other big issue. How you pay people and where you invest in marketing. These are the decisions that fall out of your attempts at sales attribution. As I've already illustrated on the first category, how you pay people, you're just setting yourself up for conflict by thinking that you can solve the attribution problem and that you can pay people individual bonuses based on individual performance.

If you'll allow me, David, I'll just jump ahead a little bit. What's the bottom line here? The bottom line here, when it comes to that first issue of compensation is I'm all for a business development person earning incentives on top of a salary. I think the base salary should be high. I'm all for it, but those incentives would be based on team performance. Performance that does not encourage the rugged individualist behavior, but performance that encourages the team to work together.

David: If we do that though, we're losing one of the most powerful beloved tools in our management toolbox. That's that we don't know how to manage these salespeople so let's just put an incentive in place that self-manages. Basically, they'll start losing weight. Finally, they'll go unconscious and then we'll have the janitor drag them out the door. We don't have any idea how to manage these people. Plus, we don't want to be held responsible for the marketing that delivers the leads they're supposed to work. We just set it up. You're basically taking away that tool. Thank you very much.

Blair: That's such a great insight. The idea that we view new business people is different so we agree to pay them differently and then because we pay them differently, we manage them differently.

David: We feel like we don't even need to manage them.

Blair: Yes. We don't need to. Hey, they're not selling, they're not earning. That's fine. That's management.

David: They didn't do much anyway. Now we have this juicy contract and I want to say, listen, you didn't do everything. It's just a little bit crazy and taking the salesperson side here, it's so disheartening to do what you think are the right things and then to say, "Hey, you didn't push it right over the line so you don't get it." One of the most demoting things for salespeople is when the rules are not super clear or they change.

I'm really with you on this. I wasn't years ago. I don't know what the percentage should be. I feel like having an incentive on top of a really healthy base is something that salespeople expect and so we don't want to demotivate them, but they shouldn't have to sell a couple of kids if they don't hit the targets within the timeframe.

Blair: We did that episode on managing tension and you were certainly pointed out that not all tension needs to be resolved. I've mentioned it a few times since then. This just struck me as a really profound idea. I think in that episode, I said this is one of those areas of tension. The salesperson wants to be paid. They want not necessarily a highly leveraged compensation plan, but they want the right to earn more. They don't really want to know exactly what they're going to earn next year.

From my point of view, you really want to pay a good person full salary and no incentives. The listener might react to that think, "No, I don't." Well, the reason you don't is because people in that role have underperformed, which might have nothing to do with that person and everything to do with you or other things going on in the firm. Without any judgment, historically, that's a difficult position.

People often perform for reasons that are theirs and reasons that are spread across the agency and you want to mitigate your risk by paying these people a smaller base and more upside. Once you get a really, really good person, you want to mitigate the downside, your downside. You want to pay them full salary. This is attention that almost always come up. The best salespeople push for incentives, but you shouldn't pay your salespeople based on individual incentives. You should pay them a high base and then possibly some team performance-based incentives.

That tension, it doesn't really need to be resolved. It's okay from my point of view, and you've converted me on this, David. If that tension remains, and there's always a little bit of tugging and neither party is really happy and you end up in this compromise that's okay.

David: How do you translate this over to the account management side? They're already a client, we've been talking mainly about bringing a client in for the first time so they're already a client. Typically, the salespeople are different from the account people. What's your perspective on the account people's side?

Blair: Well, I think in some firms, you've got a new business person who actually does have some responsibility to grow existing accounts, but in the typical firm, that's not usually the case. You get to larger and larger firms, you get to these larger dedicated account teams where you can often have large teams dedicated to just one account, the same principle applies. I don't think there should be incentives for individual performance. I think it just sets up the conditions for a civil war. Incentives based on team performances is okay.

Once the size of the firm grows and you have these dedicated teams for individual accounts or clients, then I think it's okay to start to incentivize those smaller teams rather than the firm-wide team. In the typical firm, that's listening to this and independent marketing firm of some kind, incentives should probably be based on firm-wide performance. I think the same principles apply in account management. You don't see incentive-based compensation structures nearly as often, but there's still a lot of them out there and the conditions apply.

David: Switching hats here, let's say you're the consumer, you're not the agency, do you feel different when you are interfacing with an AM for a professional service org whose compensation is in part dependent on how well the firm does? Are you okay with that as a consumer?

Blair: Yes, it's a good question. You think he is the firm's client?

David: Yes. Like you would go to, what's the big electronic store? Best Buy. Years ago, they decided to take everybody off commission and that was supposedly meant so that now whenever you got advice, it wasn't tainted. I still have a deep suspicion that they are still compensated on their ability to talk me into a maintenance agreement. That just feels real but anyway, it's just an interesting thing to note. I've had people tell me that they don't want their account people to be compensated that way, but I don't think they necessarily are thinking about a team-wide one which probably they'd be more comfortable with. I was just curious about how that relates to the psychology of selling.

Blair: I haven't thought too deeply about it from the client's point of view here, but as you point out, we've all been in situations where it's clear that this salesperson incentives are aligned to talking us into something rather than to helping us. I think most of the cases with most firms who are offering performance-based incentives to account people, those incentives aren't that high that they would cause people's behavior to go from decent to suspect.

David: Right. It might be. If they do really well, they might get a party and a song and a pizza thing or something. They're not going to buy a house.

Blair: We're back to the fundamental attribute attribution era here, why do we have this image of the sleazy salesperson? It's not because sales attracts sleazy people, it's because the incentives for them to do whatever they have to do to make a sale. That's a fully leveraged full commission salesperson. That's an achy feeling being on the buying end of that equation. Just ask yourself, are you creating those conditions with the compensation plan for your Biz Dev people on your account people?

David: Just taking a up-down, a drone view on all of this, I'm just thinking, realistically, as entrepreneurs, we're on a 100% commission system, but we figured out that we can't let our behavior be so influenced by the need to meet payroll that we compromise what we know we look for in a client and being able to live with the people that work with us on the team.

You come to this healthy balance. Really, the folks who want deep sales attribution from the sales team, they're the ones that need to explain to us why that's the case in just that situation. When the rest of us are trying to balance all the different inputs and how to tie those to the outputs. It's not as clear as we would like to think of it in a sales attribution setting.

Blair: I agree. All of this begs the question around management, which we made some mocking comments about. One of the questions is, "Well, if you're not paying a person based on incentives, how do you know how well an individual is doing?" As you alluded to it, that's a question that's really only asked to Biz Dev people. It's only asked to the highly incentivized position. It's actually a ridiculous question. How do you know how well an individual is doing?

If the listener took my advice, they would pay their Biz Dev people, a high salary, and then possibly an incentive based on firm-wide sales. It begs the question, "Okay, we've had a good year. The firm has sold a lot, but I don't know how much of it was the responsibility of that individual. How do I know I'm not overpaying them?" and the answer is, you know, you know how well an individual is doing.

In an independent marketing communications firm where you don't have hundreds of clients, you have a dozen clients. Maybe a two dozen clients. You know how well that person is doing. It's like the line about pornography. I know it when I see it. You can go through the list of all the people on your team and you can rate them A, B, C players, right?

David: Yes.

Blair: You have a sense of people's contribution even though you can't quantify it. Just let go of the need to have to quantify this position, and then the corollary question is well, then how do you keep from overpaying poor performers on a well-performing team? The answer there is you need to let go of that too. You need to accept the fact that in the short term if you're giving people incentives based on firm-wide performance, and this one person isn't contributing to firm-wide performance, that in the short term, you're going to overpay underperforming people.

That's okay in the short term. It's not okay in the long term. Back to this question of, well, you know. You know people who are contributing and who aren't contributing. If you find yourself paying incentives to somebody who isn't pulling their weight, what do you do? You let them go. You replace them. You go get somebody better.

David: All of this stems from the lack of comfort we have with the whole process. We aren't sure what kind of person to hire. We don't know how to screen them very well. We don't know how to equip them. We don't know how to incentivize them. Then when we're stuck with a situation where it's not working as intended, we're looking for some more definitive answer that we can just point to instead of facing in that uncomfortable realization that you screwed up in terms of managing this whole process. I feel the same way about timesheets, by the way. Let's say there are six people that answer to you. If you really need them to fill out timesheets to figure out- [chuckles]

Blair: Yes, Freudian slip.

[laughter]

David: -something else is going on. You should have some pretty good instincts without that level of data. It's just interesting how we scurry from discomfort by running to data when we should be thinking more carefully. It's not like data is bad. It's just data isn't all the answer especially in a sales attribution setting.

Blair: I totally agree. I spend a lot of time telling our clients in training situations that you are all good conversationalists. You know how to communicate. You know how to have a conversation, but there's something about the sales context when you're having a conversation in the sales context, when you forget most of what you know and you have this idea that you need to show up differently, and the real basics of just human communication start to evade you.

I often say if I could do just one thing for you, it would just be to reach into you and to pull out this idea that a sales conversation is different from any other conversation. You just get rid of that idea and just have you as who you are is good enough. Just show up and have a goddamn conversation which is the way we think we need to pay people. It's a similarly ridiculous idea. There's just something about selling that makes the normal rules of conversation go out the window, and the normal rules of compensation go out the window.

David: It's crazy and you think about you entering a sales situation where you were a little bit nervous about it, but it was a big result. If you landed this, it was great. Think about the person that saw you discouraged that day, and said something to you and really cheered you up and gave you some internal fortitude, or the person that came up with the great turn of phrase is going to go in the little one-page proposal or whatever.

You're telling me that's not as valuable as the person who-- It's just all crazy. We lose our minds when it comes to sales just like what you were saying. We come up with all kinds of weird incentives, and ways of doing things. I guess your sales training practice is really trying to turn people back into humans. It's not creating weird super salespeople. It's telling them to forget most of what they learned and be humans again.

Blair: Yes, and learn from various different sources not just officially trained that way. Just adopted this idea somehow from various places that the salesperson has to be slick, has to have all of the answers, can't stumble, et cetera. Needs to be talking a lot. We just let go of that idea and let go of the idea that we need to pay these people differently too. Bottom line here, sales attribution is really, really messy. I'm not saying don't try. I'm saying don't make the mistake of making some fundamental decisions like how you pay people on something as messy as sales attribution. Just accept that it's messy, and don't make big decisions based on it.

David: Yes. Everybody listening, most of these folks have a salesperson on staff of some kind or somebody that's helping. You have an instinct right now and that instinct is telling you that they need to go or they're even better than you thought. Hang on to that instinct. Sleep on it for a few days and then act on it, but don't think too much about it. Just go with your instinct.

Blair: Yes, great way to wrap.

David: All right. Thank you, Blair.

Blair: Thanks, David.