We "Insight" Success!
Joined September 2009
- Tweets 914
- Following 1,047
- Followers 387
- Likes 77
2 Photos and videos
Analysights retweeted
The wineries that win aren’t the ones that react to cancellations – they’re the ones that see them coming. By paying attention to how members behave, how they join, and where they are in their lifecycle, you can identify risk early and take action when it matters most.
#wineindustry #wineclub
wineindustryadvisor.com/2026…
2
3
3
40
Analysights retweeted
May 7
High sales during promotions don’t automatically mean customers are price sensitive.
New post:
How DTC brands can estimate true price elasticity using log-log regression, pricing analytics, and customer demand modeling.
zurl.co/SukOv
#PriceElasticity #DTC #Ecommerce #Analytics
1
2
2
27
Analysights retweeted
May 1
If customers only buy when you discount, your promotions may be reshaping behavior more than growing demand.
New post on:
• price elasticity
• customer conditioning
• and measuring promotion dependence in DTC brands
zurl.co/70xKY
#DTC #CPG #Ecommerce
2
2
20
Analysights retweeted
Apr 30
A promotion spike doesn’t always mean true growth.
Sometimes customers are just learning to wait for the next sale.
My latest post explores:
• price elasticity
• promotion dependence
• pull-forward demand
• and why constant discounting can weaken DTC brands over time
zurl.co/GFU9h
#DTC #Ecommerce #PricingStrategy #MarketingAnalytics
2
2
8
Analysights retweeted
Apr 19
Customers already tell you what to sell next.
It’s in your order data.
Most brands just don’t use it.
zurl.co/ngyO0
#ecommerce #AOV #DTC
2
2
82
Analysights retweeted
Apr 17
Most “frequently bought together” data goes unused.
That’s missed revenue.
Use it to:
→ bundle products
→ drive upsells
→ increase AOV
Start simple.
zurl.co/DippD
#ecommerce #DTC #AOV
2
2
124
Analysights retweeted
Apr 15
Most brands chase more traffic.
The better move?
Increase value per order.
Market basket analysis shows you:
→ what customers already buy together
→ what to bundle, upsell, or recommend
Simple data → real revenue.
zurl.co/BGyP4
#ecommerce #DTC #AOV #upselling
2
2
117
Analysights retweeted
Apr 11
Early churn signals:
→ Smaller orders
→ Longer gaps
→ Skipped shipments
These aren’t random.
They’re warnings.
Here’s how to use them:
zurl.co/vPbo5
#Ecommerce #Retention #DTC
2
2
22
Analysights retweeted
Apr 10
This is where most brands get churn wrong:
They react after customers leave.
The best brands predict the drop-off window and intervene before it happens.
zurl.co/V8sZe
#Churn #Growth #DTC
2
2
24
Analysights retweeted
Apr 8
The real question isn’t:
“Will this customer churn?”
It’s:
“When does churn risk spike?”
That’s when you act.
zurl.co/Stz4x
#Retention #Analytics #DTC
2
2
18
Analysights retweeted
Apr 8
Most churn doesn’t happen suddenly.
It builds over time—until it’s too late.
Here’s how to predict it before customers leave 👇
zurl.co/ZLmSd
#DTC #Churn #Retention
2
2
25
Analysights retweeted
Apr 2
We tested this:
Remove behavioral data → model performance collapsed.
Lesson:
Customer value is driven by behavior, not demographics.
More here:
👉 zurl.co/6eqzE
#DTC #CustomerData #LTV #MachineLearning
2
2
13
Analysights retweeted
Mar 28
If you’re optimizing paid media at the national level only…
You might be optimizing to a number that doesn’t exist.
Geo-level modeling changes the game.
zurl.co/nOoiP
#MarketingAnalytics #DTC #Growth
2
2
18
Analysights retweeted
Mar 27
National averages can hide poor regional performance.
A channel that looks strong overall may be underperforming in key markets.
Geo-level modeling helps fix that.
zurl.co/edbxO
#DataScience #Marketing #ROAS #MMM
3
3
21
Analysights retweeted
Mar 25
Most brands optimize paid media based on national averages.
But performance isn’t uniform.
Geo-level modeling reveals what works where—so you stop wasting spend in underperforming regions.
zurl.co/FohV7
#MarketingAnalytics #PaidMedia #MMM #DTC
2
3
36
Analysights retweeted
Mar 20
Most brands don’t have a measurement problem.
They have an allocation problem.
Media mix modeling helps solve it.
👇
zurl.co/BvPZ7
#DTC #MarketingAnalytics #MediaMixModeling
2
2
7
Analysights retweeted
Mar 18
Attribution ≠ optimization.
If you want to improve performance—not just measure it—you need media mix modeling.
Here’s how MMM helps DTC brands allocate budgets smarter 👇
zurl.co/OlTpO
#MarketingAnalytics #DTC #ROAS #MediaMixModeling
2
2
16
Analysights retweeted
Mar 13
Instead of targeting broad audiences, start with your best customers.
Lookalike modeling identifies prospects who resemble them using predictive analytics.
Here’s how the technique helps brands acquire better customers.
zurl.co/qHJjp
#GrowthMarketing #CustomerAcquisition #DataScience #DTC
2
2
17
Analysights retweeted
Mar 11
Your best customers share patterns.
Lookalike modeling uses those patterns to find new prospects who resemble them—improving customer acquisition efficiency.
In this article, I explain how the technique works and why the seed audience matters so much.
zurl.co/B9vsS
#MarketingAnalytics #CustomerAcquisition #PredictiveModeling #DTC
2
2
14
Analysights retweeted
Mar 6
Conversion rate tells you what happened.
Conversion probability tells you what to do next.
Here’s how DTC brands use logistic regression to allocate marketing spend more intelligently:
zurl.co/XWR9y
#DTC #MarketingAnalytics #PredictiveAnalytics #EcommerceGrowth
2
2
15