Highlights from a heated CMO discussion on attribution

We knew that giving the topic of attribution in B2B sales and marketing only an hour (technically two hours across two separate sessions) during last week's CMO Coffee Talk was pushing it.

Turns out not only did we barely scratch the surface, but we also triggered a wide variety of emotions relative to the complexity, elusiveness and understanding of what "attribution" does and should mean.

Below are highlights from the chat in both sessions Friday, which will give you an interesting feel for the sentiments involved!

And if you are in the CMO Coffee Talk community, don't miss Bonnie Crater 's attribution research and trends presentation in the #swipefile channel.

If you are a B2B CMO or head of marketing and want to join a community of ~2,100 of your peers, let me know and learn more here.

------------------------------------------------------------------------------------------------

Attribution hurts my brain...

“I have stupid people (in sales) trying to argue with me about attribution.”

Is it driving a wedge b/w Sales & Marketing

Should be called "attributionish"

It is the holy grail we can never really get to.....

And I see marketing leaders spend so much time trying to figure it out… that they lose focus on the things that drive business results

No shade - this still makes me chuckle: https://cdn.shopify.com/s/files/1/0573/6710/1627/products/white-glossy-mug-11oz-handle-on-left-60dcb0961497d_1024x1024@2x.png

liquor and guessing!

Its all about snapshot data!

I get frustrated that first touch, last touch, multi touch and recency to active opportunity all have pros and cons. I’ve seen them all used and all at the same time, but then it feels like juggling answers vs a specific answer. But I feel the same about investing research and stats.

Doesn’t this assume that every “campaign” object has a budget associated with it? There’s a big dif between attribution modeling and ROI.

I think about it as a combo of $$ and effort.?Some things take more "people" time (in and outside you org) that is harder to precisely measure but real.

How do you think about that reporting in an account-based multi-touch world.?I’ve found things get squishy…but generally ok squishy as long as we’re building pipe in target accounts.

attribution was pitched as the panacea?- but reality is its just another set of inisghts.

I do a QBR and we marry the results with the money

I actually stop all sourcing discussions in the company, and only allow the CRO & myself to have them behind closed doors. Of course I do attribution - first touch, last touch, and multi-touch - and by lead and opportunity source - but only the CRO & I use it as 1 piece of data. And we don't proliferate/share it. We shut the conversation down and just say that the GTM team wins together or loses together. It's stops the finger pointing and allows us to use the attribution data in the fashion it was intended, e.g. 1 tool in the toolbox in addition to others that helps us make good spend decisions!

And as marketers we don’t need convincing that “perception is reality” - if the customer “thinks” this is how they heard about us, they’re not wrong :-)

I had a conversation with the head of sales in my new gig this week and she blamed missing her numbers on marketing…when asked about rep production she has 3 who haven’t sold $1 this year…ummm- look in your backyard first!!!!!??Told her to let people go so I could have more money to build brand!

It’s not about who did what (we all win together). But you can get good insight into which touches go together to produce higher conversions.

great point - precision vs perception

We added self-reported attribution a few months ago and it provided some good insights. For example, it showed people were coming to us from Reddit so we’ve increased our focus on Reddit and have seen positive results. But it doesn’t replace traditional attribution.

We have a salesforce attribution dash with a drop down for stage. So we can look at early, mid, late — and see where they’re sourced (first/last).

that’s healthy behavior. Hard for some CMOs who don’t have the social capital with their ceo or cro who are trying to blame them and single them out. But it is the interplay between sales and marketing

Board is also an issue.?They want an easy answer.

Programs build up like a ‘plaque’. I love that!!

Job to be done: helping us steer the car, find things to cut/double down on… that’s it. If it’s about credit, you’ve got bigger issues.

Credit issues are typically about milestones,?as in acquisitions.?Milestones for engineering. and revenue, yet without support internally, they cannot be met.

The goal of 10x value of a program is interesting. I wonder if that is the expectation of other marketers.?It feels bold. But why not?

I like the self-reported attribution as one data point because - yes it’s reliant on memory - but knowing what sticks with/was memorable for your prospects is really valuable, even if it wasn’t the “first” thing they encountered.

I’ve seen really fake results on self reported. The first item on the form gets 70% or the customer identification (people don’t answer correctly- just try to get past the screen) you can randomly rotate answers to distribute it, but I don’t trust it still because of the first mentioned behavior)

There’s an inverse relationship between revenue attainment and conversations on source.

10X is where you start then every year you find the average and that becomes the hurdle

Agreed on self-reported.?The customer is "always right" (again, perception) but does their answer really push you in the right direction of where to invest programs/resources if they were wrong?

Agree! I think open text and optional is the best way to ask, completely agree with you

for orgs moving away from the sourcing model, are you still assigning revenue targets per team?

Esp as you get bigger you need to be more efficient

B2B it normally comes from multiple marketing and sales touches… so complex… there’s a “low tech” way to begin to look directionally

I use an open text field on a late stage form and wade through the data semi-manually to prevent the bias of multi-select. Also ask the SDRs/CSMs to ask the question.

Add in a partner channel and it becomes more complex.

Look at all the marketing and sales touches for sampling of won and lost deals and try to see differences in touches along the way….. the “Brute force” method...

I look at EVERYTHING- split between marketing, sales, bdr, and partner generated. And we look at first touch, last touch, and mulit-touch. And, we also look at lead source and opportunity source. So we look at ALL OF IT. We just don't dwell/focus on it! And I keep the discussion primarily with the CRO so that the teams don't argue with each on who did what! but yes - attribution is SOOOOOOO valuable to me- and I couldn't live without it. it's just how you use it.

The ultimate reality - if we’re hitting the ARR / Pipeline numbers then no one is asking about Attribution.?Marketing still should but that’s the reality…

There are always dollars available if marketing can add revenue in a reasonable payback period / 1 year or less for many boards. There is more cynicism on longer LTV than shorter

https://www.demandrevenue.com/cmos-and-csos-working-together-to-scale-marketing-and-sales/

Don’t you think the question is just as hard as accurately predicting the growth you can drive this year with a flat budget - how much of forecasting is licking the thumb 15-20% above what you can see in the data?!

“Talk to me like I’m stupid; it can’t be that complicated—it’s marketing."

Instagram and social attribution…. Wondering why our lead flow is down this week from twitter source

9 month sales cycle but of course it came down to the "one thing"....

Often there is more effort put into the attribution effort then there is in the actual campaign effort…. It can really drag you down if you have limited resources

There is no statistically viable way to say “this account came to us because of this one thing"

Mainly because Life and Business are multivariable

The cookie-less world may free us all from attribution all in the end…. ??

The different touch approaches do not account for the time tagged effects of marketing

My .02:?keep attribution to marketing convo’s, not leadership or above

I remain convinced that no one agrees on the problem statement we are trying to solve.

More specifically, the touch approach will always undervalue brand v demand

Oh and damn CPL - dear god. Let's just double down on organic...('cause that's the least expensive)...sure 'cause that is so easy...

I mean the meta question, not the how to make attribution work question

For board credibility, being 100% aligned with your sales leader counterpart is critical!

I love what you just said…if the problem is top of funnel, use first touch. If it’s velocity, last touch. Did I get that right?

There's also the problem of the total is greater than the sum of the parts. None of my campaigns show notable ROI individually, and yet we have no outbound Sales - everything comes from Marketing or Partners. For that reason, I keep most of this to myself and my team unless I need to raise a red flag. Luckily I have a CEO that mostly stays out of my business.

Behavioral Economics tells us that prospects will not know the sub-conscious impact of the influence

I think the utility of win/loss and pattern recognitions is even higher when you have a good sample size of customers. A bit harder at earlier stage companies

In addition to asking the question on the inbound form “how did you hear about us” we asked sales to clarify this early in the sales cycle and modify/clarify the answer

we keep looking for the "one thing" or silver bullet.?In complex sales situations, that's just not reality.

Especially in Business purchases. Buyers think they are making "rational" choices, when in fact their subconscious is driving the bus

My CEO wants to know the “one thing” and most of my reporting time is spent explaining that this is not possible to know

That's very true, the world is a multivariable place and any model that is not relational and dynamic is a problem.

The god damn "one thing" question - so triggered

Actually, we’ve been getting good answers to our how did you hear about us Q. At the very least, we’re finding out that it’s different from what’s reported in hubspot

Full Circle, Proof Analytics, Falkon, Bizible are the most prominent I've seen.