01 / AI + marketing technology
The question isn’t which platform. It’s what it should make possible.
Marketing technology, data and AI work that starts with what the business needs to be able to do, and ends with a team that can actually do it.
02 / The problem
Buying software is not the same as gaining a capability.
The pattern is familiar. A platform gets selected. An implementation runs. The project closes. Eighteen months later, the business is doing roughly what it did before, using more expensive tools to do it.
The software usually works. What’s missing is everything around it.
Nobody owns the data definitions. The integration between commerce and CRM was designed to launch, not to operate. The team was trained once. Reporting depends on manual work or institutional knowledge that eventually walks out the door.
This is why platform evaluations tend to start with the wrong question. Which system to buy is a downstream decision.
The upstream question is what the business should be able to do that it cannot do today, and what has to change across technology, data, process and people to make that capability stick.
03 / The frame
Five things technology should make possible.
01 / Understand the customer better
The same person can visit the tasting room, join the club, buy online and purchase the wine through retail.
The business sees fragments of that relationship across commerce, CRM, POS, email, reservations and, where available, retail data.
Not every interaction can be reconciled to one person. Better identity, cleaner data and more connected systems can still create a materially better customer view.
The objective isn’t a mythical perfect customer record. It’s enough connected information to make better decisions and create better experiences.
02 / Act without unnecessary manual effort
Renewals. Winbacks. Allocation offers. Abandoned carts. Post-visit follow-up. Reporting.
Repeatable work should run reliably without depending on someone remembering to run it.
Automation is most useful when it removes repetitive work while keeping judgment where judgment matters.
03 / Decide from numbers people trust
A member. A customer. A conversion. A channel. Revenue.
Simple words become surprisingly complicated when different systems define them differently.
Technology should help create shared definitions, clear ownership and reporting that can survive being questioned in a leadership meeting.
04 / Move faster without lowering the standard
Product information. Release materials. Email variants. Landing pages. Analysis. Internal documentation.
AI and automation can reduce the time teams spend producing, organizing and interpreting routine work.
The standard should not become “AI made it faster.” It should be “the team can do more valuable work without making the work worse.”
05 / Manage risk by design
Privacy consent. Accessibility. Permissions. Data governance. Alcohol-specific compliance requirements.
These are easier and less expensive to address as requirements than as remediation.
Good technology decisions account for governance and risk before implementation, not after someone raises a concern.
04 / On AI
Useful before impressive.
The most valuable AI use cases are not necessarily the ones that make the best demo.
They are the ones where the inputs are trustworthy, the task is sufficiently understood, the risk is appropriate and the output can be evaluated.
Today, that can include content variation, structured product information, data cleanup and classification, analysis and summarization, internal knowledge retrieval, customer-service support, and faster production of routine marketing work.
Other applications require more caution. Personalization built on incomplete customer data inherits the weaknesses of that data. Automation applied to a broken process makes the broken process run faster. Customer-facing output without sufficient review can introduce factual, regulatory or brand risk.
And wine has something particularly worth protecting: specificity.
Estate. Vineyard. Vintage. Appellation. Farming. Winemaking. People. Place.
AI should make it easier to organize, retrieve and communicate what is distinctive about a wine business, not average that distinctiveness into generic luxury language.
Search is increasingly mediated by AI systems that extract, combine and answer from information across the web. Clear site architecture, structured data, consistent entities and factually rich content therefore matter beyond traditional search optimization.
Wine businesses have unusually good raw material for this. Provenance, appellation, varietal, vintage, production and place are inherently structured facts. Many winery websites simply don’t expose those facts particularly well.
The practical position: start with useful problems, improve the data underneath them, test what creates measurable value, and keep human judgment where the stakes require it.
05 / Adoption
Implementations rarely fail on software alone.
Technology projects succeed or fail in the operating model around them.
Who owns the platform after implementation? Who owns the data? Do marketing and IT agree on what the important fields mean? Were the people expected to use the system involved in designing how it works? Is someone responsible for improving it after launch?
This work is less visible than platform selection. It is also where much of the value gets created or lost.
My experience here comes from two very different environments.
I’ve owned a global marketing platform inside an enterprise IT organization, where governance, integration, security, privacy and cross-functional alignment are unavoidable.
I’ve also led digital and marketing technology inside wine businesses, where teams are smaller, resources are constrained and the same person may be responsible for strategy, implementation and the revenue result.
The scale changes. The requirement doesn’t: technology has to work for the organization that actually has to use it.
06 / The work
Where I usually start.
Martech assessment
What you own. What it costs. What it does. What overlaps. What nobody uses.
And which problems are actually technology problems.
Capability roadmap
Start with what the business needs to be able to do, then work backward to the technology, data, process and organizational changes required to get there.
Sequenced to the team and budget you actually have.
Platform selection + migration
Requirements definition, independent platform evaluation, vendor process and implementation planning.
The objective is not to select the platform with the longest feature list. It is to choose the system the business can realistically operate and grow into.
Integration + data
How commerce, CRM, email, POS, analytics and other customer systems connect.
Where identity can reasonably be resolved, where it cannot, which system owns which data and what needs to move between them.
AI use cases
Identify the work worth augmenting or automating, evaluate the quality of the underlying data, assess risk, establish success criteria and pilot before scaling.
Governance + enablement
Ownership models, data definitions, documentation, training and operating practices designed to keep the capability working after implementation.
07 / Experience
Both ends of the scale.
I’ve worked with marketing technology as both the business owner asking it to produce a result and the technology owner responsible for making it work.
Schneider Electric
Technical Product Owner, Salesforce Marketing Cloud
Owned roadmap prioritization for Salesforce Marketing Cloud inside a global enterprise IT organization, working across marketing, technology, data, cybersecurity, privacy and regional business teams.
The work included platform governance, marketing enablement, integrations and translating business requirements into an enterprise technology roadmap. At that scale, governance isn’t theoretical. Neither is adoption.
Crimson Wine Group
Director, Digital & Revenue Marketing
Led digital and marketing technology across seven wine brands, including implementation of Salesforce Marketing Cloud for 120,000+ subscribers.
The work crossed ecommerce, CRM, personalization, analytics, accessibility and GDPR/CCPA privacy requirements, with different brands and teams operating at different levels of digital maturity.
J Vineyards & Winery
CRM & Digital Marketing Manager
Led migration to Salesforce and built segmentation and lifecycle programs on top of the new customer infrastructure.
The objective wasn’t the migration itself. It was what the marketing team could do differently once it was complete.
Advisory + transformation work
Today I work across DTC, marketing and IT on ecommerce and digital transformation, including platform modernization, marketing technology decisions and customer journey design.
That perspective is useful because technology decisions rarely belong entirely to marketing or entirely to IT. The important ones sit between them.
08 / When this makes sense
The usual moment.
A platform decision is imminent.
The vendor conversations have started before the business requirements are clear.
You’re paying for capabilities nobody uses.
The technology stack has grown faster than adoption.
Marketing and IT disagree about who owns the customer data.
The organizational problem is becoming a technology problem.
Leadership is asking what you’re doing about AI.
You need a considered answer that is more useful than a list of tools.
Every important report requires manual work.
The systems contain the information, but the operating model around the data doesn’t work.
The implementation launched. The capability didn’t.
The technology is live, but the business is still operating much as it did before.
Or there is an important technology decision coming and you want someone who can evaluate it from both the business and platform side.
09 / Start a conversation
Have a technology decision worth thinking through?
If the question sits somewhere between wine, customer experience, data, AI and technology, let’s talk.