I very often get asked what is the ideal customer for Digital Alchemy and the answer has been the same for many years. We like to work with customers struggling to get value out of their marketing automation investment. The majority of our work is remediation and helping customers get to where they thought they were going to get to when they implemented. We speak regularly with organisations who have gone through the pain of implementing a platform like Salesforce, Braze, SAS or Adobe but are unable to get the value from the implementation they were expecting or committed to.

Here are 5 things to focus on in 2025 that will help you break through:

1. It’s All About the DATA

One of my biggest gripes is implementations built based on an MVP framework or designed around a small number of use cases. Why is this a problem? It’s a problem because if you ask a vendor to build you a Mini they will build you a Mini, and it will work perfectly well as a Mini when you want to transport 4 people, but you can’t turn a Mini into a bus that transports 50 people. That’s when the pain starts. You quickly realise that strapping people to the roof of your Mini and stuffing people in the boot is then shuttling backwards and forwards like 10 times is not a solution. 

So if you need a bus to deliver value to your business, don’t start with a Mini design. When we look at clients in this situation, most of the issues are with the way the data infrastructure has been designed. All of the platforms that I mentioned above work — they all do the job, maybe they do it in different ways and they have different strengths but they all work. The issue is 90% of the data.

• You can’t run real time campaigns if you don’t have all the data in real time
• You can’t automate a campaign unless all the data required for the campaign is delivered consistently 100% of the time
• You can’t automate a campaign if there is any manual data processing before the campaign executes
• You can’t build automated campaigns unless you understand the data

Focus on the data design, and make sure your data is designed for automation.

2. Have a Vision, Not a Plan

Building a high performing automated marketing system is not a 3-month project. You must have a vision for the customer experience that you want to create, and how you want to be interacting with your customers in 3 years’ time. Without this vision, how will you know what capabilities to invest in, what team to build and what partners you need to engage to help you?

I’ve seen companies buy real time marketing when they don’t have real time data, online targeting when online sales are 5% of total sales, and implement a new platform with no plan to migrate existing campaigns. 

Your technology and your data need to support your customer vision, not the other way around.

3. Think About the Customer

Customer centric marketing is so old that it’s new again…but seriousl, it is so much easier to take a campaign centric approach but here is a chilling example from a company I saw when we did an audit. Over a year, 30% of their customers were opting out, how I hear you gasp… well maths never lie. They were blasting their customers between once and twice a week, blasting as in everyone who had not yet opted out. If you look at the individual campaigns only 0.5% of customers were opting out of each campaign, that’s not a terrible result considering the revenue from each campaign. However, if you sum that up over 60 campaigns per year…it adds up to 30% of the base.  How did they not realise this, easy, it was masked by customer acquisition, the base was still growing slightly because they were acquiring more than they lost. Cutting the annual opt out rate from 30% to 10% increases the database growth rate by 20%…that’s just one example of how a campaign vs customer approach can yield very different results.

4. Build Quality, Then Scale

Hyperpersonalisation does occur by accident, and it certainly doesn’t happen easily.  To get there you need a higher level of quality in your processes than even for regular campaign automation.  I see many attempts of hyperpersonalisation that work well in limited controlled environments but fail to scale.  If you are going to build a hyper personalisation POC then design it so that it can scale. Concepts only make money if they can be executed at scale.

5. Innovation is a Job, It’s Not a Sideline

Operations always trump innovation.  The most common story I hear from frustrated marketers is that they can’t innovate and experiment because they are overwhelmed with BAU. Constantly waiting for a gap in the BAU so you can innovate is like waiting for snow to fall in the Sahara, it happens but not very often and when it does it doesn’t last long enough to ski on.  If you want/need to innovate then it has to be a role or at least carved out as a distinct responsibility.

If you want to build an automated marketing system you need to design it that way, you can build manual and ad hoc processes and then run like crazy…but that’s not marketing automation that’s marketing exhaustion.  New implementations are sexy, but they deliver no business value, value only comes when you operate at scale.  So in 2025 focus on operating at scale, and find a partner who has a track record for doing that 🙂

---

by Regan Yan, the CEO of Digital Alchemy.

Regan is a subject-matter expert in analytical database marketing and customer relationship marketing, as well as an in-demand presenter and keynote speaker at national and international events. He also authors thought leadership pieces on data-driven marketing that can be found on the DA Blog.