Here’s what I learned this week

Tare.ai is cool.

I made a post about it on Friday, I have had multiple requests for calls to discuss other AI tools around landing pages and a few more around email and ads.

Everyone is working on this problem.

One of my contentions is that any SaaS ecommerce company’s silent goal is to replace agencies and transfer that $3,000-$10,000 per month into software that does 85% of what an agency can do for between $500-$2000 per month.

We’re about to see a gold rush of these kinds of companies.

I think it’s one of many that will popup until the big ESPs decide to release their internal versions, the thing is for someone like us, it’s the data payload that all these services are lacking. The process is still too manual.

And that data payload has to come with context - guess what we don’t push into other platforms? Yup, the data context is our moat.

You want to be able to feed data directly into the platform and have it spit out ready made suggestions based on the data.

We’re back to this needing data thing aren’t we…

Which brings us to the topic of the week -

Topic of the week: Data Interpretation

Last week I talked about algorithms and data.

This week I want to go through a bit more about where the gaps lie in understanding and interpreting data.

This is a real screenshot from a real popup that we use on our website.

It’s sandwiched between a few other questions and is only shown on our landing pages which is where we drive ad traffic to. Targeting matters here, remember that.

Brandon Bobart asked the right question.

This is where a lot of people struggle with data and data collection, they ask the questions, pick out answers, but they don’t fully understand how to leverage the data collected or the answers provided.

So here’s a breakdown of how I would leverage the answers to the above question, these were on the post but I was limited by space, so this is the expanded version.

Comments on Ads

If people are commenting on your ads it means you made them stop for long enough to take an action relevant to something that you created. Engage with people who’s attention you’ve captured, algos love attention and engagement, if it’s coming from your existing audience hire someone if you need to that response to all the comments.

You need help, just write ChatGPT with this prompt: “Respond to this comment, [insert comment], in a playful and joking manner in the voice of [insert actor name].”

This is your one chance to interact in the vibe of your brand outside of customer support, seriously spend some serious time doing this.

Product Images

If people are telling you that product images made them click and that subscription to conversion rate based on that answer is high, then it’s time to go product picture first on both your ads, including slides of all your product shots on your product pages, maybe some carousels, and go into details of explaining those product pictures on your landing page.

I don’t think I’ve ever seen a landing page that actually dissects all the photos that people put in their product pages up close with descriptions.

(I’m adding this to a landing page creation project today)

We’re visual people, so the more images of something the better.

Offer

If someone responses with offer, we need to make sure that offer is front and center on the ad, shows early, and that there is continuity on the landing page. We can’t hide the ball with an offer. Hit people hard right away.

Re-enforce the offer throughout your copy and your landing page.

Pure Curiosity

If pure curiosity is winning out then it’s time to ramp up your charm. If you’re good enough to get a click and a subscription, you’re good enough to close this deal, they’ve already taken multiple acts of intent.

Saw an ad, Clicked an ad, Subscribed to a newsletter, you just need to lay it on thick and give them that old razzle dazzle relevant to doing good things.

Alex talking about underwear

This is kind of a random one thrown in there, but there’s a lot of content where it’s Alex the founder talking about the underwear, we put him front and center in a lot of ads and more. So this is just trying to understand how many people saw him prior and didn’t take action or are taking action again.

I’m in the market for undies

This question seems obvious, but it’s not, it’s a good litmus test to see the kind of audience that Facebook is serving our ads up to and it helps us adjust copy and imagery to talk directly to this audience both in the ad and on the landing page.

“It looks like you’re in the market for an underwear upgrade”

Ad copy can now be about what to look for in your next pair of underwear, how not all underwear is created equal, why any feature and benefits are things you should consider for your next underwear etc.

Here’s how ChatGPT did when I prompted it with the same above information in case you don’t have one of me around

All in all not bad. Some of it is vague and basic, but overall this is a good starting point.

Every answer has the possibility of being leveraged in the hands of a capable marketer, of course without context, it’s hard to build strategy.

Now it’s a little bit of a bitch to have to type all that out without context of revenue, orders, and conversion rates on which things to prioritize and focus on, if only there was a way to dynamically populate all those prompts on the back end and just produce reports with context for people to take action by leveraging data…

Oops, wait a second, some of us have both data and context…

Then what if those reports could be used to create secondary prompts to be fed into other systems to create automatic emails or ads for brands leveraging existing brand assets and creative short punchy copy overlays…

If only there was a company that already had built the platform to collect and analyze all this data and produce results that could be easily plugged into other programs capable of creating generative ai that were not only backed by just user feedback but user feedback directly tied to sales and revenue…

I haven’t been shy about telling you that AI is coming for all our jobs

I’ve been saying this because unlike a company that has to find the data, build the prompts and then create new things with them to take action, we already have the data, already have the analysis already, have the prompts, now it’s just a matter of connecting to the services that are capable of ingesting those prompts and turning them into collateral.

I really see a lot of ecommerce being automated in the next 12 months.

Email - The lowest hanging fruit

Ads - Pretty Straight Forward with the right templated structure

Landing Pages - Pretty Straight Forward with the right template

Product Pages - Pretty Straight Forward with the right template

THERE’S A GAP THOUGH…

The one hole for all of these things is having reliable data to know what’s working. I’ve been telling people for years now, the only thing that matters is quality of audience.

The only way to understand if you have a quality audience is to understand what makes an audience quality and more likely to convert than others.

I’ve begun tracking a new stat around this…

First time Orders broken down by subscriber or non-subscriber.

It’s trending at 40% of first time orders are coming from people subscribing for a discount through one of our forms and providing data.

This means that these models above can be built on the data of 40% of all first time purchasers which is massive.

You’ve heard me say that I don’t care about SMS before, this is the reason, it’s absolutely useless as a value to use in building for the bigger picture.

Give me data all day long that I can leverage to build these other models, rather than watching 26% of interested people drop off and fail to provide me with anything that’s actionable.

One day people will start to understand how this stuff works, my hope is that in reading this you get a better idea of what’s required to make this stuff work for real, not some fake pixel stuff, but actual business strategy coupled with data and tools to make stuff work.

There’s a reason why all these content sites are signing massive deals with AI companies, you need data sets to build this stuff.

We just happen to have a really large intent data set with contextual relationships.

This is why I am pretty confident that AI will be much faster than most people think in terms of taking over parts of the industry.

The Takeaway

Ecommerce is tough, it’s getting tougher, it’s going to be really interesting to see what happens over the next few years.

Have a great week!

-Jon

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

You can learn from me: jonivanco.com