Shivan Durbal

Start with probing questions that when strung together paint a picture as to why the hesitation, and then follow that with active listening. Sometimes there are completely rational reasons for the hesitation brought on by their history with trying or not needing it before. Change is hard. Through this you can form a plan to help unstick your client's reluctance which may include a well managed and risk managed test design to help them ease into it and measure effectively.

The mainstays for automotive locally I've found are Google, Meta and Local Ads/Directories. What sets you apart as an Automotive advertisers is Creative and Messaging and how you use Carousels if available, Lead Ads, Conversation Ads and find new ways to engage customers prior to the test drive. Media/Channels alone aren't enough - the best Automotive companies that are over indexing to local advertising are doing 'conversation marketing', and doing it extremely well. Furthermore, they are also spending time in, and around, their community building 'less than high fidelity' content to build up that ground swell and conversation.

A few things I found helpful when I started looking was to create a few variations of my resume to target slightly different roles I was qualified and interested in, and then went to work. I quickly learned a good bio also helped people know me better. Then I posted about an orbit of connected topics I felt uniquely able to weigh in on or that I could answer at the top 10% of LinkedIn talent. Further, I lit up a free version of canva and got very good, very quickly, at short story telling. It's served me well now, and back then. It worked so well in fact, I met fantastic people along the way and we're building a technology to replace bios and read me's. A link.tree for professionals!

AI and ML powered insights are critically important, and never more accessible. But accessible doesn't always mean legible! Designing the best systems that do the most good, should also include stakeholder level data visualizations that empower the user, within those stakeholder groups to do a few things exceedingly well - with little oversight. Those are: Access Data, View/Toggle that Data, Derive Insights Confidently, and Then Take Action...all without having a background in Analytics. Do this well, and you can move an entire organization, or at least many departments towards the same Marketing (dare I say Business) Objectives and Goals. The 'good stuff' happens when teams dig in together, and create the new, NEW, thing!

In my view, the overlooked pillar in designing and then integrating ML into a digital strategy (or most strategies), is Use Case Design. Inevitably the model, assuming it does as it's intended to do, will be put into production to help carry out tasks and improve Revenue Driving, Cost Saving, Experience Enhancing, or Capability Creating outcomes to name a few. The best models I've seen that 'do the most good' for an organization, rather than becoming a science experiment, are those led by data scientists and teams that have a commercial or human (or ideally both) defined outcome in mind. Do this first - then yes, get the data quality right, ingestion, unification, etc etc. But first, get that right, and the rest will fall into place.

I break out AI into buckets to make sense of it all. Today vs. Tomorrow. Today - use it as a prosthetic, to enhance daily workflow such as tasks you do everyday, insights you need repetitively, information better gleaned by ML/AI or to super power an original idea with multiple derivatives. Future State - keep an eye on LLMs, 'Action Models', co-pilots you can speak to plainly and which return plain responses rooted in vast data sets, and how ML/AI can help offset what we will inevitably lose in re: to audience transparency when cookies go away - there are other ways to predict and target, using other bid stream signals. In fact, keep an eye out for how the space is using different signals to replace 'old ways' of doing things.