Four Automations Were Published, Green And Error Free, And None Of Them Could Ever Run

Key takeaways

• On September 2 we checked the reference build of a CRM template we sell and found four automations that were published, showing green, with zero errors, and could never run once.
• One of them was a 26 step nurture sequence.
• The template is built in two halves on purpose. Small automations apply tags. Portable campaigns start when a tag appears. That split creates one failure the platform never reports: a campaign waiting for a tag nothing ever applies.
• Not every tag with no listener is a bug. Some are applied by a nightly script, some by a person following written steps, and about twenty mark a finished journey on purpose.
• The fix is a proof script that checks every campaign's trigger against everything that could apply it, with a short list of known outside sources. Each entry has to say who applies the tag.

As a GHL consultant in the Bay Area, I get asked to look at a lot of automation setups. Most owners can show me a long list of workflows, all published, all green. What almost nobody can tell me is which of them have actually run.

On September 2 we asked that question about our own product, and four of the answers were never.

How the template is built

One of the things we sell is a CRM template: a full build you can set up in your own account, with pipelines, forms, follow ups and campaigns already wired. We keep a reference build of it, the version we treat as the product.

It's built in two halves, and that's deliberate.

The first half is a set of small automations that know about one specific account. When a form gets submitted, or a deal moves to a certain stage, they apply a tag to the contact. That's all they do.

The second half is the campaigns. The nurture sequences, the onboarding steps, the check ins. Each campaign starts when a certain tag shows up on a contact, and it doesn't care how the tag got there.

That split is what lets the campaigns move cleanly between accounts. The account specific stuff stays in the small taggers, and the campaigns travel. It's the right design. It also opens one gap the platform has no way to show you.

Green, published, and never going to fire

If a campaign starts on a tag, and nothing anywhere ever applies that tag, the campaign just waits, forever. It's published and it shows green, with no errors, because nothing has gone wrong. Nothing has happened at all.

We found four of those in the reference build. A partner onboarding sequence. A check in for partners who'd gone quiet. A referral step. And a 26 step nurture sequence, all of it written, timed and published, waiting on a tag no automation, script or person was ever going to put on anyone.

Not every quiet tag is a bug

The first version of a check like this would be simple: list every tag a campaign starts on, list every tag an automation applies, and flag anything on the first list that isn't on the second.

That check would cry wolf constantly, because there are legitimate ways a tag gets applied that the platform can't see.

Some come from a script that runs every night outside the CRM, reviews the book of business, and tags contacts that need a look. Some get applied by hand, by a person following written steps in a task. And about twenty tags work the other way around. They're applied on purpose and nothing listens for them, because they just mark that someone finished a journey, so they can show up in a saved list later.

So "nothing applies this tag" and "this is broken" are two different findings, and a useful check has to tell them apart.

What we built

We built a proof script for the template. Among other checks, it reads every campaign, collects every tag a campaign starts on, and checks each one against every automation that applies tags. Anything left over gets checked against a short list of known outside sources.

The rule for that list is the part I like most. Each entry has to name who applies the tag: the nightly script, or the person, and in what situation. The code comment says it plainly. "Something applies it" is not a verifiable claim.

Any tag that isn't applied by an automation and isn't on that list fails the check, by name, every time it runs. It says which tag, which campaign is waiting on it, and that nothing applies it. We run it before we claim the build works.

What this means for Bay Area businesses working with a GHL consultant

If you bought a snapshot, a template or a done for you setup, you probably own some automations that have never run. Usually nobody did anything wrong. "Published" and "can actually start" are different things, and the dashboard only shows you the first one.

The question to ask about any automation you're counting on is short. What has to happen for this to start? And has that ever actually happened?

Three things worth doing this week:

• Pick the three automations you'd be most upset to lose. Open each one and read its trigger. Write down, in plain words, what has to happen for it to start.
• For each of those, look at how many contacts have ever entered it. Zero after months is a finding, not a slow week.
• If a trigger depends on a tag, find what applies that tag. If the answer is "someone's supposed to add it," ask who, and whether they know.

Any GHL consultant in the Bay Area should be able to answer that question for every automation they hand you, not just show you a screen full of green.

Want to build this yourself?

Join the On Point Tech Academy at optechsol.llc/academy. It's free, and we build live every Tuesday and Thursday.