Here’s what I learned this week

Seems like everyone is talking about the need to collect zero party data.

Also people are talking about quality of their lists.

The second party I’m not sure they have a good idea of how to measure.

So today we’re going to talk about Quality of Audience.

What it means, how to measure it, and what you can do today to get a jump on the rest of your competitors.

The Exchange that kicked off this edition of the newsletter

Abhishek Patra posted the following:

I have audited over 300+ ecommerce brands.

This is the no.1 issue that holds them back 👇


They are focusing on the wrong things.

Brands focus primarily on (in order of priority):
▸ Design
▸ Copy
▸ Strategy
▸ Attribution
▸ Deliverability

But, the email and SMS list should be your top focus. 🔑

A strong list with average emails will generate more revenue than a weak list with great emails.

So, you need to focus on:

▸ Improving the quality of the list

▸ Gathering more information on the list

▸ Increasing the list size using the current list

▸ Keeping the list engaged for as long as possible

▸ Increasing AOV and repeat purchase rate of the list

▸ Decrease the time between purchases for the list members

Then, figure out the strategy, copy, design, segmentation, personalization, and AB testing that will work for the list.

In the world of first-principle thinking, the priority needs to be on the email and SMS list.


I won’t go through all of this, but a lot of this is fluff.

You can only improve the quality of a list by improving the source of the traffic and the messaging that is relevant to your end goal, purchase.

Increasing AOV and repeat purchase of a list is something we’ve covered before, you can’t influence retention as much as you think, it’s up to the perceived value of the product, use of the product, need for more of the product, and timing.

Meaning the only thing you can really influence is perceived value or price, the rest are variables up to the end consumer that are incredibly hard to influence due to their individual behaviors.

Same goes for decreasing time between purchases, you only have one lever to pull.

But this was the question and answer string worth paying attention to:

Jon: “"Improving the quality of the list" what do you do to know if the list you have is quality? How are you currently measuring this?”

Abhishek Parta: “🏌🏼‍♂️Jon Ivanco For a brand, by quality we mean - how many people end up buying once and more than once and what makes them buy - discounts or value.”

Jon: “Abhishek Patra 🛡️ so you don't know if they are quality before they purchase?”

Abhishek:”🏌🏼‍♂️Jon Ivanco no. Zero party data can help but no sure shot way. Know anything?”

Jon: “Abhishek Patra 🛡️ been using zero party data to make all my ad buys for years based on data combinations with context.

If you base your quality on Purchase timing, you win.

There are other statistical combinations that can get you there depending on the brand, but we've found that purchase timing is the leading indicator of an audience being quality relative to orders, revenue, and conversion rate.”

Then something interesting happened…

Alexandre: “Abhishek Patra 🛡️ I don't even think we need to complicate things with 0-party data or anything.

0-party data are there to get us more data and help qualify leads for the right product

But if the leads coming from the ad suck. 0 party data won't do anything”

Jon: “Alexandre DA COSTA you're probably correct, with how you're currently looking at zero party data, not how you SHOULD be looking at zero party data.

Here's the difference...

Your zero party data doesn't come with context to help you on the acquisition side. I'm betting the zero party data you collect is only being maybe leveraged in email flows.

This is a huge gap if you're only sending in values rather than actually contextually understanding the relationships between the values and things like orders, revenue, and conversion rates.

I'd argue you really aren't using zero party data correctly or in a manner than can have an outsized impact.

Here's a video you should watch: https://youtu.be/LaDEcmTEoKE”

Alexandre: “🏌🏼‍♂️Jon Ivanco list quality stems from your message & offer on the ad. That's the first touch point driving people in your list, then the page congruency with the ad. I think the pop up helps convert that's all. That's my view on that.

I think best way to measure it indeed is purchase. At the end of the day if people in your list don't buy you can assume the quality isn't there if your offer are compelling.”

And this is the magic moment…

Jon: ”Alexandre DA COSTA how do you improve the offer and message on the ad via data and statistics without collecting zero party data?

Should we just guess? Or should we maybe leverage contextual data to build a full funnel marketing plan that is statistically more likely to convert?

I'm pretty sure you've just defined the reason for collecting and leveraging zero party data, go give that video a watch.”

Circular Logic

So this is a great example of circular logic, rather than focus on using tools to improve the relationship between data and strategy, we have someone literally defining the reason they should be using data, but arguing against using data.

It doesn’t make logical sense.

I see this a lot too, where people like to hold true to their guns around certain things and pass the blame on another silo.

This is a great example of why email who owns the popup should either relinquish control to the acquisition team or start to actually logically admit to the role that a popup plays, one of the two.

So let’s dive into this stuff, because it seems like a lesson worth teaching today.

How we leverage popups to figure out quality of audience and why you should too

Popups aren't about opt-in rate.

They should be about data collection that leads to global marketing strategy.

If you're not leveraging them for that, you're just pissing away margin in search of an immediate upside.

I've read multiple posts lately about quality of audience with no real way of measuring whether an audience is quality or not unless people purchase.

Responses tend to be, did they purchase, and did they purchase again.

There's no way to tell the difference in a sign up outside of them taking the final action, which means there's no way for you to influence that potential action in a meaningful way.

Reminder that people that purchase from signing up via your popups do it really fast.

So then the focus switched to opt-in rate.

It's a numbers game they say.

If you're gaming for opt-in rate, just do this:

1. Full Page
2. Right Away
3. Move the close icon
4. Question First
5. Great Offer

Bonus would be just run a giveaway, you'll get a good sign up rate.

If you do this though, you'll typically see a decrease in subscription to conversion rate, you'll giveaway more margin than is potentially necessary, and you'll likely still only get an email and maybe a phone number.

I'm going to propose an alternative that we've been using for years.

Ask for data relevant to the customer journey.

There's one key question you should be asking.

When are you looking to purchase?
Today
In a few days
In a few weeks
In a few months

You should see a natural correlation with the answers and results, by and large there is a predictor in the timing of the purchase relevant to conversion of the subscriber.

So logically speaking, when I run acquisition campaigns I base the success of an ad campaign or spend based on the amount of signups I get that correlate to looking to purchase Today.

This typically outperforms every other ROAS based calculation for one important reason. (Remember ROAS can't be measured 1:1 in a timeframe)

More people purchase the first time, more people have the potential to purchase a second time.

It's all about product in hands.

Now only 63% of my first time sales come from subscribers.

Again numbers. so of my total subscriber pool of opt-ins on our popups I know that roughly 60% of my subscribers are looking to purchase today and that around 45% of those will go on to make a purchase.

I like those odds.

They also stay consistent for long periods of time.

So when someone asks me what a quality audience is on my list, I look to zero party data with contextual sales data to tell me what makes up a quality list.

But Jon, I know why people are purchasing, they have pains that they are trying to overcome.

You’d be surprised, a lot of people buy things just because they are trying new things, they might have something they are trying to solve for

What does quality of audience mean?

Quite often I hear that a quality of audience is one that purchases.

That’s a customer not an audience.

One of the main things that has struck me over the years is that with the current system people are using, there is no strategy for understanding the difference between subscribers in ecommerce through popups.

We don’t treat customers differently based on the form they signed up on, nor do we typically collect the proper amounts of data to determine their intent.

This is the top question that will determine the quality of your audience.

When is someone looking to purchase.

The correlation of answers always will sway to people that are looking to purchase today.

Time and time again, we’ve seen that ads that drive more signups with the answer “Today” perform better.

Which makes sense.

So when we look at the performance of an ad for instance, we know that if it has a good signup rate and people are answering “Today” the audience has a higher percentage of chance of converting.

We can look at the other answers and assign a probability to the overall performance of the ads that we’re running to see if they are appealing to a high quality audience.

How to measure quality of your audience

So you’re now collecting this data point, what’s next? Well now you can look at all your signups week over week and start to rank your audience quality based on multiple data points.

You’re now going to be able to predictively model your revenue based on your traffic.

What? What did you just say?

So for a long time, quality of audience has been super difficult to understand because everything was tied into “does someone purchase” rather than “what’s the probabilistic outcome of someone showing intent and purchasing”.

The issue is that the best way to manufacture a purchase is to provide an incentive for someone to purchase, then collect data relevant to their current mindset around making a purchase.

This is zero party data in a nut shell, but people haven’t been leveraging it properly to understand these relationships.

So the natural progression is a forward modeling based on past statistics and traffic figures to tell you that if you drive X new customers, then statistically we expect to see, X new subscribers, and X new customers at a conversion rate based on X with an AOV of X.

Now we’re able to map this out based on the data point collected at the time of signup to match up to the existing historic data points that have statistics and probabilities baked in.

Is this exact, no of course not, but with enough data points we’ll have a solid confidence interval around the likelihood of someone purchasing.

When it comes to marketing there isn’t an exact science, there’s just data collection and interpretation that can influence decision making.

We use the above system. You should too, because it works.

How to build a system to better understand your quality of audience

I could tell you that you should just collect zero party data, put it all in a spreadsheet, run all the calculations, separate multiple forms, use variations of questions across different forms based on traffic source etc using your existing tools.

In truth though, none of them were setup to accomplish the above.

You don’t get to this level of approach with off the shelf tools, you have to build your own, challenging previous beliefs.

I talk about this stuff because I’m passionate about it. I talk about it because before we built Formtoro, none of this was remotely possible without a shit ton of manual work and then some.

Believe me I used to have structured data, then I’d have to make lots of manual pulls to get me even 1% to what we currently have in Formtoro.

And we’re only going to get better on focusing on this sort of quality of audience.

There’s a few key steps to building this system.

1. Install Formtoro

2. Read my newsletters

3. Ask questions

I really do think that the key to long term success for most brands is going to be leveraging direct data from customers to build better marketing strategies.

You’re just de-risking an industry that is seeing decreased margins and a steeper uphill battle.

How to attract a quality audience via your ads

My favorite topic, illustrated beautifully by the exchange above.

“Alexandre: “🏌🏼‍♂️Jon Ivanco list quality stems from your message & offer on the ad. “

He’s right, it does, but you need to know what messaging drives the most about of sales, most revenue, and the highest conversion rates.

And the best way to get there is to ask your customers during their highest point of intent, right before making a purchase, then structuring that corresponding data in a way that provides you the insights necessary on figuring out who you’re talking to and what matters to them in their buying experience.

This part actually isn’t tough with data.

It’s just offer and messaging.

If you have data that specifically tells you the correlations between user provided data points relative to purchases then you have everything you need to create a brief aimed at that strategy.

The link is above but I go over this in detail in:

How to maintain a high quality of audience

Attracting and maintaining a high quality of audience are kind of one in the same.

Stay the path, follow the data, message to the group that brings you in the most money.

That’s it.

The Takeaway

Quality of audience, know it, understand it, live for it.

Have a great week!

-Jon

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

You can learn from me: jonivanco.com