Multi-Touch, What am I Missing?

[Chris Leslie]

In Brian’s latest podcast he is arguing for the use of Multi Touch Funnels in stead of using the CRM to make marketing decisions.

I don’t disagree with this idea but I haven’t heard anyone talk about how to calculate assists into ROI. or maybe I am just missing something.

Let me give an example.

Lets say Autotrader costs me $3,000/mo
If AT sends me 10 leads and I close 2 at an avg of $1000
Then my ROI is -$1000 for the month.

But what if they have 14 assists?

How does that factor in to ROI?

What’s those assists worth?

If I credit AT with a dollar value then I have to remove it from somewhere else?

 

[AdamMurray]

I don't know if I add much to this conversation, but I will say this...

1. I'm thankful this is a discussion and we are making strides to move past the ridiculous notion of last-click attribution.

2. I'm old...and I hate math. I know we all strive for that magical ROI calculation that allows us to make quick, easy decisions while hiring and firing vendors and lead sources and still sleep soundly every night. However, there will never be a 100% solution to this predicament. People aren't bots and clicks aren't programmable. We should never forget the human element, in customer terms, and as marketers we must employ a healthy dose of gut instinct.

To paraphrase one of the great pseudo-sales guys of our time...the leads aren't shit, you're shit. Attribution is not binary. What was the content? What was the pricing? So many questions need to be asked before applying a simple formula (=sum(leadcost/leads)) and making a binary decision.

 

[Chris Leslie]

All I am saying is that if we are going to have a conversation about something other than last click then there needs to be some sort of value assigned to the other things. I am asking how do we do that?

So for the sake of keeping this thread focused. Lets assume that all the other questions have already been answered and are good.

 

[marcelus]

Assigning attribution and getting an idea of what went into a sale can be pretty tough the further away from the sale you get. For example: you might run facebook ads for your dealership and not see much direct traffic from them when they run, but tons of customers see them and just remember dealership XYZ exists. Months later when they're in the market for a new car they check your dealership first just because they remember it exists or they remember the name before a competitor. They may get reported as a direct conversion from your own site, but they're really a conversion from the facebook ads. Ads you might not even be running any more because of not being able to attribute many sales to them and discarding them as a waste of resources.

You may be able to get a better idea of how to weight things by surveying customers and asking how XYZ influenced them buying from you.

It would also be interesting to see the different approaches people have tried.

 

[Chris Leslie]

I probably should have posted the podcast of Brian's I am referring to. He is claiming that we shouldnt trust the sources in our CRM.

http://brianpasch.libsyn.com/stop-guessing-and-dont-trust-your-crm

 

[AdamMurray]

All I am saying is that if we are going to have a conversation about something other than last click then there needs to be some sort of value assigned to the other things. I am asking how do we do that?

So for the sake of keeping this thread focused. Lets assume that all the other questions have already been answered and are good.

Respect.

To that order then, I would think you need to assign a weight to the attribution itself. Sources will have to split deals just like our salespeople do.

 

[Steve Stauning]

I've been listening closely to those selling attribution to dealers (and those paid to sell attribution) for a very long time; and I will have more to report about this after I've gathered enough information to confidently provide guidance to my clients. (I have follow-up meetings scheduled over the next several weeks with some folks preaching that leads are basically worthless and everything is multi-touch attribution; so I want to give them a fair shake.)

That said, much of the multi-attribution argument so far sounds like this: "Well, Bob gave the guy a brochure last week at our south store and the customer showed up and bought at our north store, so Bob deserves half of the deal."

 

[Chris Leslie]

I’m really interested in hearing more about how that goes Steve.

Another question I started thinking about while looking at my example is.

If we say Autotrader has 14 assists. Are those people looking at my cars? Or are those people just on Autotrader doing whatever?

 

[ed.brooks]

Are we talking about generating leads here, or about selling cars? Sales from leads make up the minority of most dealers sales. People who walk in after doing extensive research online make up the majority of sales.

Measuring influence is now possible, so we can move beyond the 'where did this lead come from' question as being the most crucial when we look at a marketing budget.

This, from Feb of this year, goes to the heart of the 'CRM isn't that helpful when it comes to marketing decisions' issue -

Consider the scenario Steve White wrote about in this blog post,
"A customer first discovers your dealership via a third-party auto listing site. Then they come back to your site later by searching your dealer name on Google and clicking on an AdWords listing, which perhaps was right above a free Google organic listing for your website. By default, in Google Analytics, that AdWords click took 100% credit for that visitor, assuming as a fact that it was incredibly valuable, when it may have had no impact at all on whether or not that sale was made."​

So back to my original question, are we talking about generating leads here, or about selling cars?

 

[Chris Leslie]

Make it about whatever you want but I’m only talking about digital traffic And digital conversions.