- Why MarTech often falls short and how predictive AI can boost results for festive year-end/holiday campaigns.
MarTech doesn’t usually fail because the tools are bad. It fails because the practices are not in place.
In many companies, data is still scattered across POS, e-commerce, social, CRM and loyalty. The MarTech stack is layered on top of these silos, so what you get is more dashboards, not better decisions. Marketers end up using 10–20% of what they’ve bought: basic blasts, some simple segmentation, maybe a few automated journeys… but nothing truly predictive or contextual.
Festive campaigns make this worse. Year-end is noisy, deadlines are insane and everyone is chasing “reach” and discounts. So brands fall back to the same playbook: blanket 11.11 / Christmas / New Year promotions, generic retargeting, last-minute media buys and very little learning carried forward into the next season.
Predictive AI changes the game in three ways:
- Who to talk to: Models can score customers by likelihood to buy during the festive period, likelihood to churn, or likelihood to respond to a particular category or bundle. Instead of blasting everyone, you can prioritise: high value customers who are at risk of going quiet, “swing buyers” who only shop during big sales, and new customers with high lifetime potential.
- What to offer and where: Predictive models can surface product affinities, discount sensitivity and preferred channels. So your “12.12” campaign stops being just “storewide 12% off” and becomes: personalised bundles (e.g. “kids + home” vs “gadgets + travel”), smarter promo depth (save deeper discounts for those who truly need it to convert), and channel-optimised journeys (app push for some, WhatsApp or email for others).
- When to act and how much to spend: AI helps forecast demand at SKU level and even at store or city level. That means you can align inventory, staffing and ad spend to where demand is most likely to spike, instead of reacting once things are already out of stock or CPMs explode.
Practically, I always advise brands: don’t start with “AI everywhere”. Start with one or two use cases for the year end season. If you can prove a clear lift there, the appetite to scale predictive AI across the rest of the customer lifecycle goes up very quickly.
- Malaysia as ASEAN’s MarTech hub: what must government, industry and universities do to make this real?
Malaysia already has the right ambition on paper. MyDIGITAL and the Malaysia Digital Economy Blueprint explicitly position the country as a regional digital leader by 2030, with targeted initiatives around AI, cloud, data centres and digital talent. Budget announcements in recent years have also doubled down on AI and innovation as growth engines. The question now is execution, and MarTech is a very practical test of that ambition.
For the government, other than incentives and grants, we also need clear, practical guidance for marketers and tech vendors especially around consent, data portability and cross-border AI workloads so innovation doesn’t die in “legal says no”. On top of that, having more accessible and regulated “testbeds” where Malaysian and ASEAN firms can experiment with first-party data, clean rooms and predictive models for industry specific sectors like retail, healthcare and tourism will reduce compliance fear while accelerating learning.
For the industry, brands should treat data governance, experimentation and incrementality testing as core capabilities, not “nice to have”. This is the only way to prove MarTech’s value at the board level. On the other side, agencies and consultancies need to evolve from “media and content vendors” into builders of repeatable products such as CDP accelerators, attribution models, sector-specific AI engines that can scale across ASEAN, this means easy integration into regional e-commerce, payment and logistics platforms, not closed ecosystems or just a one-time content project.
For universities, the real shortage in MarTech is not just coders or creatives but the “translators” who can speak marketing, data and regulation in the same sentence. Universities can deliberately design for that. We need graduates who understand funnels, CDPs, consent strings, cloud basics and marketing analytics and not only the 4Ps or SWOT analysis. This means joint programmes across business, computer science and law. Final-year projects should be structured around real datasets and real problems from Malaysian companies. Let students build actual models and dashboards that the companies can deploy.
- How cross-border data collaboration in ASEAN can unlock better predictive models while respecting privacy.
Most ASEAN markets, including Malaysia, still train their models on relatively small, fragmented datasets. At the same time, ASEAN is negotiating the Digital Economy Framework Agreement (DEFA), which explicitly focuses on harmonising digital trade rules, including cross-border data flows and data protection. This includes federated learning, data clean rooms and anonymization of data amongst many. If we get this right, cross-border data collaboration can unlock: richer, more diverse datasets for training predictive models, better insights into cross-border customer journeys (e.g. travel, e-commerce, remittances), and more robust fraud, risk and identity models. If we can align on “privacy by design” by having strong consent, clear purposes, accountability, and privacy preserving technologies, ASEAN can enjoy the benefits of shared data insights without creating a surveillance or compliance nightmare.
- Gaps (talent, infrastructure, policy) that slow AI marketing adoption across the bloc.
From the talent perspective as above, we have a shortage of “translator” roles. There are not enough people who can sit between marketing, data, IT and legal. Many marketers are still uncomfortable with data engineering and model performance, while many data teams don’t fully understand funnel metrics or brand building. The challenge in ASEAN is also where some markets are aggressively investing in AI and analytics skills while others are still figuring out basic digital marketing. This creates a two-speed region where only certain hubs can truly leverage advanced AI. For the talented, they are often pulled into global roles, leaving local brands heavily dependent on vendors. That’s fine at the start, but over time, it weakens internal capability.
Infrastructure is a constant conversation because many companies still operate separate systems for e-commerce, CRM, loyalty, call centres and offline retail. Without a unified data layer or CDP, AI models end up being “channel-specific hacks” rather than end-to-end intelligence. I have seen something as basic as consistent product IDs, event tracking and consent capture is still missing for a lot of brands. This is outside of the fact that a huge part of ASEAN’s economy is driven by SMEs who are still on very manual workflows. Brands need to remember that AI can’t fix poor data hygiene.
As for policy, it is nice to see that ASEAN is moving quickly; but on the other hand, this speed is also uneven across borders. For example, data protection regimes are at different levels of maturity, and cross-border transfer rules are still evolving. In Malaysia, for example, PDPA amendments and new guidelines have strengthened obligations around data governance and cross-border transfers, which is positive but also raises the bar for compliance. Without more aligned standards across ASEAN, every cross-border campaign or data initiative feels like a custom legal project, which is expensive and slow.
Until we address these gaps in a coordinated way, for many marketers the safest path feels like “do nothing with AI involving personal data”. That slows experimentation and encourages a checkbox mentality rather than a risk-based, innovation-friendly mindset.
- Opportunities arising from digital transformation and the urgency for companies to adapt now to capitalise on seasonal demand.
The festivities always translates into very tangible opportunities. Companies should use this season to own the customer relationship with first-party data. Festive campaigns are a golden moment to grow logged-in users, loyalty members and opted-in audiences. The brands that use this year’s season to build their data foundations will have a structural advantage next year (if you don’t have them for this season). With the right digital stack, companies can adjust prices, bundles, media budgets and inventory allocation almost in real time. Seasonal demand becomes something you shape and not just something you react to when the warehouse is already empty.
This is especially true since customers don’t just want discounts; they want relevance. Predictive AI, combined with good journey design, lets you recognise returning customers across channels, anticipate what they’re likely to need next, and delight them with timing, content and offers that feel intuitive rather than intrusive, or just blasting the same generic information to every customer out there. Things like AI-assisted customer service, smarter logistics routing, and better workforce planning reduce friction when demand peaks can also directly affect revenue and customer satisfaction.
In conclusion, treat the upcoming season as both a revenue moment and a transformation sprint. Pick a few high impact areas for first-party data capture, implement a predictive model for a key use case, and have a more integrated view of inventory and demand. Commit to doing them well and the compounding effect over the next three to five festive cycles will be far greater than any single hero campaign.
Legal Disclaimer: The Editor provides this news content "as is," without any warranty of any kind. We disclaim all responsibility and liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. For any complaints or copyright concerns regarding this article, please contact the author mentioned above.
