Here’s what I learned this week

I spent most of this week talking to people building some cool stuff.

Data is changing the game in ecommerce.

People are doing a lot with technology in new and different ways. A lot of what used to take a long time to do research around is being done by computers in seconds.

But like I’ve talked about before, most of it lacks real context.

We don’t even evaluate reviews based on order size.

Is a review the same if it’s someone that spend $200 as someone that spent $40?

Are they a different type of buyer?

The Obsession with Attribution

I’ve long been a proponent that attribution doesn’t matter.

I take a completely contrarian approach to ecommerce digital marketing.

I only care about quality of audience, relevancy of messaging, and conversion.

I don’t think any of these things is channel dependent. I think that it’s likely that your audience exists in multiple places and as such your messaging is relevant on multiple channels too.

The problem with most attribution is that we only look at sales to determine quality of audience.

When we measure ROAS (Return on Ad Spend) in a 1:1 period we don’t seek to understand what traits make people more or less likely to become a multiple time purchaser (which is the actual goal).

There’s a difference between these two people:

1. One purchase $100 total lifetime

2. First purchase $40, but second purchase $200, and likely to make a third purchase

But when we look at ROAS on a campaign, number one is going to look better every time. This might be one of the biggest problems with all of attribution reporting.

The vast majority of brands and agencies love to copy each other and use existing frameworks and hooks. If it’s not broke don’t fix it is the mentality.

A lot of this leads to marketing that all looks the same.

So if no one has a good beat on what conversions are likely to lead to multiple purchases within the standard timeframe to make media buying decisions then is it possible that winning ads aren’t really winning ads?

I would contend that it’s not only possible but likely.

I’ve heard that a lot of brands are going for 1x ROAS as a baseline these days, it’s almost as if the goal posts moved to align with the lack of strategy and alignment.

Which brings me to something else I want to talk about, the cost of creative these days when it’s likely that none of it is being properly reported on…

I just saw some recent pricing for some creative work and it’s ridiculous.

Just a hunch, but if I was a SaaS company, this is a space I would look to disrupt knowing that no one actually knows how to do proper reporting on what’s working or not because of the fallacy of 1:1 ROAS.

Someone with software is going to do a better job with better data and be able to cut this bill by 80% more cost effectively.

When we see things like this it’s quite possible that there needs to be a different way of thinking about these things.

Why does an ad fail to hit it’s target?

This is actually the question we should be asking ourselves.

It’s usually just Audience and Creative. You can look at these.

I go into a lot more detail in the full course which has been used by a number of people to build ads that perform just by following a data driven strategy.

But we see this a lot with posts on Reddit about ecommerce too, lots of people just throwing stuff out there and hoping that things work out and algorithms do the work for them.

That’s not a strategy. Nor is creating a ton of creatives and hoping for some magic to happen.

Without a real data driven approach beyond sales, it’s just guessing and hoping.

The other problem is smaller budgets, the way that these algorithms work, is the more data they get the better they perform, most of the time.

For the most part the difference between a successful ad and a non-successful ad comes down to the messaging that is attracting the proper audience.

Ecommerce is actually a game of understanding statistics and probabilities.

This post went largely ignored last week but it’s a framework that any brand can use to better understand where people are at in the customer journey:

Marketing Approaches Relevant to the Customer Journey

You have to meet your audience where they are in relation to the product you're offering.

Here's an example of the categories for the market for Formtoro:

Category 1: No popups (not a good fit, big opportunity)
Category 2: Just email popups (good fit, big opportunity)
Category 3: Email and SMS popups (good fit, big opportunity)
Category 4: Single zero party data point popups (great fit, big opportunity)
Category 5: Quizzes and/or Post Purchase Surveys (great fit, big opportunity)
Category 6: Zero Party Data with multiple steps popups (great fit, tougher sale)

This is the loose evolution of how people get started with popups then work into data collection.

On a lot of my calls though I've realized that most people don't know the baseline stats of their own businesses.

So we need to leverage statistics to set the stage and build a baseline of understanding:

80% of customers will only buy once
50% of customers will buy without subscribing
50% of subscribers will never open an email
70% of subscribers will never purchase
75% of purchases from subscribers will come within 12 hours of signing up
40-60% of first time purchasers come from subscribers
90% of revenue from email campaigns comes from repeat purchasers


Now the stage is set, it's time for us to pull out what the goals of anything are, increase revenue, optimize CAC.

So let's break that down into plain English as a baseline.

75% of sales from subscribers will come within 12 hours of signing up.

That means that for the most part their decision to purchase was already decided before signing up. Collecting an email or phone number is something that you would have received after the purchase anyway. You just sacrificed margin to influence a sale for the same information you were going to get anyway.

50% of subscribers will never open an email.

With Apple automatically telling you that emails have been opened you'll have to realize that once someone signs up, the odds of them interacting with you brand drop by about 50%. So that email you collected might as well be used as retargeting display advertising in someone's overcrowded inbox.

Ok so what happens after 12 hours to the 50% of subscribers that are actually opening your emails?

A few things...

70% of subscribers will never purchase.

Sucks but it's true, even if they show intent the majority will not end up purchasing even after signing up for a discount. If you collect just an email there's no way for you to establish a pattern of what makes someone more or less likely to purchase, there's no context.

80% that do purchase will never do so again.

Even if someone does purchase the first time, without context as to why they were purchasing, we can't follow up with them to understand if the product met the expectations. We don't have context for the purchase with just an email.

Context is solvable, people just need to understand why it's needed.

The Takeaway

It’s Masters weekend so I’m watching golf, it’s time to rethink what we’ve been doing and how we’ve been approaching it.

Have a great week!

-Jon

Catch up on past posts: https://ecomwithjon.beehiiv.com/

You can learn from me: jonivanco.com