Are you ready for an AI-driven marketing transformation?

Marketing transformation.

Two words that sound impressive in a boardroom. Terrifying when you realise no one actually knows what they mean. Even more so when you have to justify the cost of changes and impact in an already uncertain marketing ecosystem.

Yet we find ourselves in this strange boat – the CMO, their team, the numerous agencies and everyone in between claiming to be on an AI-driven marketing transformation journey. It’s the equivalent of the Eat, Pray, Love journey of healing and being better, but I digress.

As with meditation, take a deep breath with me, pause and let’s reflect together. Before we spend six figures on a tech stack making more promises to better our customer experience, and improve conversion all sprinkled with AI keywords (as if it is the cure-all for everything), let’s ask ourselves the hard questions we’ve been avoiding in our mental CMO therapy sessions.

Transformation isn’t magic. It’s not a quick-fix. It won’t solve for problems until you decide what problems you want to solve. Most importantly, it’s about people, processes, tech and a lot of convincing the management and other organisational functions to do better.

It’s not an expedition for the faint-hearted.

Here’s my 2 fils/2 halalas on what you actually need before you can transform anything:

1. A business strategy (One that isn’t vague or ChatGPT’d)

Before we embark on the arduous task of re-organising teams, choosing methodologies, technology platforms and processes or even hiring consultants (guilty!) let’s ask: Why are we transforming?

If the answer is anything like “to be more AI focused or data-driven” or “because our CEO or national leaders said so” or “because the vibe is right”, please let’s save everyone the agony and stop immediately.

Real marketing transformation, or what we like to call AI-driven marketing reinvention, only works when it’s rooted in clear business goals. The ones with clear key performance indicators (KPIs) and objectives and key results (OKRs).

What are we trying to do?

Increase revenue
Improve customer acquisition
Improve reputation
Avoid being publicly embarrassed by competitors

Write it down.

Back it up with industry data and research.

Edit. Fine-tune.

Make a good business case. (A good-looking PowerPoint deck helps.)

Make sure the rest of the organisation agrees with you.

2. Executive buy-in (One where they all say “Yes”)

Vague excitement is not a buy-in.

Enthusiastic nods and words of encouragement are not buy-in.

The CEO and CFO saying, “as long as it doesn’t impact budgets” is not a buy-in.

As it is with consent, we need a formal, enthusiastic “Yes” and commitment for support for different stages, roles and involvement as needed during the process.

Real transformation needs active support from your peers:

The CFO, who must believe ROI is more than a fairy tale happily ever-after.
The CTO, who will inevitably say, “This is not how our stacks work” or “We need an on-premise solution”.
The CDO, who owns the data you need but is not willing to share because of ‘governance concerns’.
The CEO, who backs you and won’t change direction halfway through.

Throw a formal meeting with catering, good coffee and cakes and get everyone to be a part of the AI-driven transformation.

3. Data that’s not just aspirational (or mythical, like the Yeti)

Everyone says they’re “data-driven.” Most are just driven by dashboards they don’t understand. You don’t need all the data. Just the right data.

You’ll need:

A single view of the customer (harder to have than you think)
Clean, compliant, accessible data (zillion scattered Excel sheets don’t count)
A basic grasp of where your data lives and who actually owns it (to avoid internal political quagmires)

If your team is still manually managing data through Excel sheets, then maybe start looking into your data strategy first.

4. Talent that understands what’s needed

Transformation doesn’t just require new tech, it requires new thinking. It requires a can-do attitude and resilience through change. It requires an open attitude.

It also needs:

Marketers who understand data
Technologists who understand consumer and marketing tech
Analysts who can explain things in people-speak
Creatives who are open to “performance metrics”
Procurement that understands deliverables

But here’s the catch, hiring “unicorns” won’t solve problems.

Instead, upskilling existing teams, building team capabilities and getting them to work cross-functionally are key to success. And all this while managing the teams and ensuring they don’t over-step on each other’s roles, jobs, nerves and egos.

If your team fears change more than agency rates, that’s a red flag.

Culture beats capability.

5. A roadmap that doesn’t resemble a fantasy novel

Too many transformation plans are equal parts jargon and delusion. A clear value realisation roadmap, with ongoing measurement tracking key milestones, is way more helpful.

Something more realistic helps:

Start with small wins (email personalisation counts)
Build iteratively
Prioritise what actually moves the needle
Short-term vs long-term milestones

Remember, every quest needs a map. It’s ok to stop and ask for directions.

6. A tech stack that doesn’t need a degree in engineering

Marketing technology is fabulous when it works. Many CMOs already own more software than they can use, and adding more doesn’t necessarily fix the problem. It just makes your stack look like a pizza with too many misplaced toppings.

Before you buy anything new:

Audit what you’ve got.
Ask: does anyone actually use it?
Confirm: will this new platform solve a real problem, or just give us another login?

If your martech diagram looks like modern art you are struggling to decipher, then you are lost and need to revisit.

Transformation is mostly being very good at change management.

Marketing transformation is 10 per cent tech, 20 per cent data, and 70 per cent getting people to work together without passive-aggressively CC’ing the entire company.

It’s slow. It’s political. It’s often exhausting.

But if you’ve got: