Marketing Technology | Customer Engagement | Full Stack Marketer | Speaker/Trainer
Joined September 2010
- Tweets 2,774
- Following 590
- Followers 629
- Likes 505
184 Photos and videos
Everyone in the room knew the answer.
Nobody moved.
Knowing is safe. Doing is exposed.
Analysis is not a substitute for a decision.
1
Your martech stack has no owner.
Gartner 2025: teams use only 49% of the tools they pay for.
A tool with no owner never gets cut. It gets renewed.
Accretion beats governance, quarter after quarter.
phronetos.com/diagnostic
Governance is not a document. It is who decides.
A named owner for what enters and leaves the stack. A cadence that retires what stopped earning. A rule that every tool maps to a customer job.
If you cannot name who decides, you do not have weak governance. You have none.
1
Governance is not a document. It is who decides.
A named owner for what enters and leaves the stack. A cadence that retires what stopped earning. A rule that every tool maps to a customer job.
If you cannot name who decides, you do not have weak governance. You have none.
Only 12% of companies can prove the revenue their AI actually generated.
The other 88% are spending on faith. The difference is not the tools.
The 12% set the metric before they buy. One outcome. Holdouts. Lift, not clicks.
1
45% of marketing leaders say their AI agents already failed them.
The model isn't the problem.
82% know clean data must come before AI. Only 26% can enforce it (LeanData, 2026).
Your agent runs through that gap at machine speed.
Buy the model last.
3
Your churn model is 90% accurate and still the wrong tool.
Accuracy tells you who is leaving. Not who you can save.
A churn score treats a persuadable and a lost cause the same.
Stop scoring who leaves. Start funding who you can keep.
2
You cannot buy orchestration. Every vendor will sell it to you anyway.
It is the decision about which tool owns which job and how they hand off. That is architecture, and it belongs to you.
Orchestration is a responsibility you keep, not a feature you purchase.
5
You cannot buy orchestration. Every vendor will sell it to you anyway.
It is the decision about which tool owns which job and how they hand off. That is architecture, and it belongs to you.
Orchestration is a responsibility you keep, not a feature you purchase.
5
Your highest-scoring lifecycle test reaches 800 people.
The one you keep skipping reaches 90,000.
ICE scoring has no reach term, so it sorts your roadmap by cleverness, not value.
RICE adds reach. In lifecycle, reach is the whole game.
1
You pay for a martech stack you use less than half of.
Gartner: only 49% of tools get used. Nobody owns the other half.
I built Phronetos to find the leak. Free diagnostic gives you a governance score, 0 to 100.
phronetos.com/diagnostic
1
35
Your attribution dashboard has never measured one incremental dollar.
It splits credit for conversions you already won. It cannot tell you which would have happened anyway.
Attribution tells you who to thank. A holdout tells you who to pay.
10
You do not have a martech problem. You have a seams problem.
The tools work. What fails is the space between them.
Consolidating onto one suite just moves the seams inside one vendor.
Unify the seams, not the platform.
2
You do not have a martech problem. You have a seams problem.
The tools work. What fails is the space between them.
Consolidating onto one suite just moves the seams inside one vendor.
Unify the seams, not the platform.
1
A dashboard full of green metrics is not the same as a business that is working.
I trusted numbers that were accurate and misleading for years.
Not because anyone lied. Because we measured the wrong things well.
Five questions I ask before I trust any dashboard.
You do not have a martech problem. You have a seams problem.
The tools work. What fails is the space between them, where data hands off and ownership blurs.
Consolidating onto one suite just moves the seams inside one vendor.
Unify the seams, not the platform.
1
The best strategic move I ever made looked like I did nothing.
A competitor launched. Everyone wanted to match immediately.
I pushed back. We waited. The launch stalled. The signal was noise.
Urgency is easy to manufacture. Patience requires a stronger argument.
1
OpenAI opened self-serve ChatGPT ads in May 2026.
Most teams will spend on the new channel before they can measure it.
A learning agenda comes before a spend agenda. What you want to learn determines what you fund.
What did your last test actually teach you?
16
You pay for a martech stack you use less than half of.
Gartner: only 49% of tools get used. Nobody owns the other half.
I built Phronetos to find the leak. Free diagnostic gives you a governance score, 0 to 100.
phronetos.com/diagnostic
1
4
Every 2026 GTM deck says: go ecosystem-led.
HubSpot's partner channel, now 100M dollar plus, was built by its 15th employee against orders. His pitch was rejected twice.
You don't plan an ecosystem. You notice the rep already building one, and you don't kill it.
7