December 8, 2025
Max Kreynin
Why Our Industry Needs a Real ComparativeTest**
(With AI, Digital Inventory, Cubesheets — and Real-World Results)
_________________________________________________________________
In December 2025, a LinkedIn post from Zach Rattner, CEO and Co-Founder of Yembo, sparked an unexpected chain reactionacross the moving industry.
He announced that their new AI vision model improved survey accuracy by 30%:
“Yembo just released a new vision AI model…improving accuracy by more than 30%.”
https://www.linkedin.com/posts/zachrattner_yembo-just-released-a-new-vision-ai-model-activity-7399431801464786944-IT2v
A bold statement — and a meaningful one for anyone involved in pre-move inspections and volume estimates.
Because if accuracy improved by 30%, it implies that:
In practical terms, this means a potential 40% variance in a field where every cubic meter matters.
And that triggered a simple, long-standing question:
How can any system — human or AI — determine packing materials and box counts based solely on exterior views of closed cabinets?
This question had been circulating privately for years.
This time, it landed publicly.
_________________________________________________________________
Here was the question:
How does an AI survey determine the number of packing boxes required for the contents of a closed cabinet — when it only sees the outside?
To Zach Rattner’s credit, he didn’t delete the comment or ignore it.
He responded honestly:
“We make an assessment based on millions of historical inspections, and allow the mover to configure behavior based on their preferences.”
That transparency deserves respect.
And it inspired a bigger idea.
_________________________________________________________________
(AI surveysvs. digital inventory vs. cubesheets — judged by actual operational results)
What if we finally answer the question the entire industry has been dancing around?
Let’s run a real test.
A test that compares:
Then, compare all predictions with actual field results:
📦 actual inventory list
🚚 actual truck and container load volume
📦 actual box and material consumption
No marketing claims.
No theoretical accuracy.
Just reality.
A survey tool is only as good as its abilityto predict:
Let’s measure those outcomes directly.
_________________________________________________________________
To eliminate bias, respected industry leaderscould act as neutral referees:
People who have no need to embellish or protect egos — only the truth matters.
This would be the moving industry’s version of the old circus wrestling tradition known as:
“The Hamburg Score”
When performers gathered behind closed doors in Hamburg to determine who was actually the strongest — no audience, no promotion, no illusions.
So let’s ask:
Who’s ready to go to Hamburg?
_________________________________________________________________
(The real challenge is box count — and the consequences are massive)
As several commenters pointed out, including Max Kreynin:
“Packed volume estimate is one of the core products of any survey tool.”
True. But the real operational pain point is not cubic volume.
It’s box count.
Let’s take the same cabinet.
From the outside:
But what’s inside changes the move entirely:
_________________________________________________________________
Scenario A: Clothing
Scenario B: Books
The cabinet is identical externally —
but operational requirements differ by a factor of three in box count and seven in weight.
If the estimator does not explicitly indicate “book boxes”, the crew will compensate by using dish-pack boxes.
Each dish-pack with books weighs ~70 kg.
At that point:
The estimator may want to avoid meeting the crew for a few days.
This isn’t a theoretical problem —
it happens in the field every week.
And this is precisely why AI-only exterior recognition cannot reliably determine packing material requirements.
_________________________________________________________________
The LinkedIn post generated immediate interest:
https://www.linkedin.com/feed/update/urn:li:ugcPost:7402668451577966592
Some of the responses:
“Brilliant idea… rather than continuing with placebos and guesswork, you can orchestrate a real test.”
— Max Kreynin, Managing Partner, Voxme Software Inc.
Rumor has it that someone in the industry has already started taking bets on how the test will unfold.
A Hamburg score indeed.
_________________________________________________________________
This is not about ego, or about proving one tool better than another.
This is about:
Whoever can accurately predict:
…will define the next decade of moving technology.
_________________________________________________________________
The conversation started with a simplequestion:
How can you know what’s inside, by looking only at the outside?
That question has now evolved into an opportunity:
A neutral, transparent, real-world comparative test of survey technologies.
The Hamburg Score of the moving industry.
We’ll keep you updated as events unfold —
and as the contenders step into the ring.
Head office:
28 Finch Avenue West, Unit 201,Toronto, Ontario, Canada, M2N 2G7
