It’s a new year again! And lots going on both in and outside the industry. That whole “may you live in interesting times” phrase is apocryphal but we are in for an interesting year. As a result, we’ve been thinking at Tactic Lab about how to simplify our agenda for 2026, as well as our recommendations for clients, to focus on the bigger picture. Here’s what we came up with – we consider these our top 3 things we’re keeping in mind but they could be part of your top 3 as well.

1. Without a good data foundation, you’ve got nothing

Ad blockers, data breach, privacy, free speech – we didn’t start the fire…

With all the things impacting your data, you’ll almost certainly need more/better data than you have. So you should take every opportunity to make the absolute most of what you have. It’s well beyond the list below but some things to get you started:

• Are you tracking user interactions with your brand across your digital platforms, or are using multiple services/tools that form part of the user journey that are untracked?
• Are you saving data that will expire (eg. Google Analytics, Google Search Console) somewhere like BigQuery?
• Do you have an optimised CRM? Are you reporting based on it? Are you linking that to your other data sets?
• Do you have your data consolidated somewhere where it’s easy to access/combine? These days it’s not just enterprise-level organisations that can benefit from a data warehouse/lake – it can be beneficial (and not expensive!) even for small orgs.

2. Spend time on channels where you’re less likely to be screwed over

We sometimes talk about owned vs rented channels and that’s still a useful frame but it’s not a strict binary. It’s more of a scale. Below are some channels starting from the ones where you’re least likely to be screwed over:

• Your website/app – if people already want to visit/download it, you have most control over that. The only way that can be stopped is if a government or ISP blocks it or the app store blocks it.
• Your CRM – I suppose you could get your account suspended for being in breach of something but that’s very unlikely.
• Your email/sms platform/strategy – you do need domain sending reputation and these days senders need to jump through more hoops to ensure that services like Gmail deliver their emails. But once someone is subscribed to you, you still have some of the best access to them possible – so make the most of it.
• Your business listings (eg. on platforms like Google) – these have their quirks and algorithms too but are very incentivised to serve accurate information about your org when people ask for it. This could even apply your SEO for branded keywords or what responses LLMs give people who ask about your org.
• Information about your org in systems that respond to user input (eg. non branded SEO) – here, you’re very susceptible to algorithmic changes and you can expect these to experience “interesting times” this year.
• Information about your org in systems that are algorithm-first (eg. social media) – here, you’re not just susceptible to the algorithm but that’s all that exists. It will not get easier to be seen, especially since you’re implicitly competing with everyone’s current content in the firehose that is the social feed.

We think you should be realistic with where you allocate your time and factor in risk. The more time/effort you place in the channels towards the end of your list the riskier. It might be unavoidable for your industry but you should still make sure you’re investing in the lower-hanging fruit.

3. Testing over trust

Every man and his dog is now AI-first. You shouldn’t be surprised if your supermarket-bought potato peeler is AI-first, IoT enabled, and forces you to sync with the cloud. This means you are being bombarded with tools you use to run every part of your organisation promising you that things are different now, you have access to real intelligence that will do things better than you or your staff/suppliers ever could.

Is that true? Maybe. Maybe not. We recommend you put things to the test and find out. The same goes to work being done by people of course, but we’ve been especially mindful of “AI-optimised” promises that some of our clients have seen that haven’t panned out. If you don’t run a test, you probably won’t know.

You can be cynical and watch the scene in The Invention of Lying where Mark Bellison first learns that you can say something that isn’t true but we can be more generous these days. You can ask yourself: if the marketing copy promising that a tool does a certain thing was written by an LLM that hallucinated, would you find out?