When I take on a new email automations client, I usually spend the first 30 days getting to know the account.

Spoiler alert: Flow logic means you can achieve the same thing in a LOT of ways. Figuring out why a client did something prior to my getting there sometimes takes a bit of time, but it’s worth it. Today, we’re cracking into Hey Bud’s account.

Anyway, I start by building out a custom flow map for the client that looks like this when you zoom out:

While I’m mapping, I’m spot-checking emails, flow logic, filters, etc. Here’s a list of some of the low hanging fruit I find that immediately improves the account:

• 🍏 Incorrect filters blocking email delivery

• 🍎 “Skip recently emailed profiles” enabled (usually I don’t like to have this on for automated emails - I would rather skip a campaign than skip the triggered emails based upon behavior)

• 🍐 Dynamic sections setup improperly

  • Eg. Almost all of you have an Abandoned Checkout flow

    • Most of you call it an Abandoned Cart flow

      • This is wrong

    • Most, when they find out they’re missing the Abandoned Cart flow, duplicate the Checkout flow and use the Cart trigger but dynamic events for Checkout don’t work with Cart

    • Most don’t have the right events setup in those Cart emails because they’re using the Checkout events

• 🍊 Emails saying they’ll deliver a code but they don’t

  • This could be because they setup the dynamic code wrong

  • Or they just forgot to include it in the email

• 🍋 Buttons that go nowhere because there’s no link

  • Happens more often than you’d think

This is just to name a few. But! The issue I recently found took the cake and it wasn’t a user-error… it was a Klaviyo error.

And it behooves each of you to scour your account for this, because when we fixed it we jacked the Welcome Flow revenue by 2.3X.

THE ERROR: CONDITIONAL SPLITS

If you’re not familiar with how these things work, you ought to have a path leading off in two directions from a conditional split. In this case, the split was a 50/50 randomized traffic split. Instead, it had one path leading to an email, and even though for the time period in the screenshot it said that 1,247 profiles either passed through the split or were waiting in it, the following email showed ZERO recipients.

If this were at the top of the flow, it’s likely that the Hey Bud team would have discovered it because the flow would have generated Sweet F.A. in terms of revenue. That’s pretty noticeable.

Instead it was, quite insidiously, at the bottom.

In a welcome flow, two emails make the most revenue:

1. The first 👆️ 

2. The last 👇️ 

This error meant that none of the latter emails were being sent. Klaviyo disputes this. But, here’s why they’re wrong:

WHY KLAVIYO IS WRONG

Prior to fixing the automation (which required a full flow rebuild) the Welcome Flow was the 7th highest revenue earner in the flows. 🚩🚩🚩 This is super-duper uncommon. Normally, it’s #1 or at least in the Top 3.

I’m blurring everything relevant for privacy, but this is a screenshot of the 6 months prior to enacting the fix. Languishing down in #7 is the Welcome Flow. And before you notice that it’s on Draft, that’s because it’s on Draft now not during that time period. 👌 

Post-fix? Up in the Top 3. 💪 

THE RESULTS

Okay, so what’s the difference between these two and how do we find it? Bearing in mind that the new welcome flow has had limited time to run. (About 30 days as of right now...) Still, we can use it as a baseline.

When you consider:

1. How much revenue per day the original generated over its last 6 months of operation

2. How much revenue per day the new flow generated over the past 30 days

…the per-day result is a 2.3x improvement.

This is why I think Big K is wrong. Not gonna tell you how much revenue exactly, but play this game with yourself: Take the revenue driven by your Welcome Flow for the past 30 days… then multiply it by 2.3.

And that was in the first month of working with me. Hadn’t even sent my first invoice. Very satisfyingly, this fix paid for over 12 months of my services. 💎 

Finding these errors is both insanely gratifying (because I know I’m adding value immediately) but also a bit heartbreaking. This thing “broke” four months prior by the looks of it.

How much revenue was lost? Too much. Back on track now? Yes, yes, y’all. 💪 

WHAT CAN YOU DO:

1. Spot check all your flows or at least the highest earners

2. Hire me and I’ll do it for you

Not kidding, I usually pay for myself well in advance with the tests I run and faults I correct.