Should You Fix Your AI Trust Signals Yourself, Or Hire Someone To Do It For You? An Honest Answer.
- Patrick Moorhead
- Mar 18
- 5 min read
When Marcus speaks, he gets some version of the same question almost every time someone corners him after a keynote.
"Marcus, we need to fix our AI trust signals... fast. Can you help us? Where do we even start?"
I've watched him give the same honest answer consistently: it depends. But not on what most people think it depends on.
Before he answers the how, he always makes sure the person understands the what and the why. Because the how only matters if the stakes are clear.
First: Let's talk about what's actually happening to your business right now.
The issue is this: AI models, ChatGPT, Gemini, Perplexity, and the others your buyers are using every day, are making active decisions about which brands to recommend and which to filter out. They aren't neutral. They aren't just surfacing whoever ranks on Google. They are running your brand through a set of trust criteria, and if you don't meet the threshold, you don't get mentioned. Your competitor does.
This isn't a future risk. Across the 5,000-plus companies we've analyzed in the AITS platform, 89% are failing or scoring low on Authoritative Outbound Citations. 78% are failing on Advanced Schema. 76% don't appear on Google Page 1 for their core topic. These aren't vanity metrics. They are the specific reasons AI models are declining to recommend those brands right now, today, in conversations with their potential buyers.
The impact is what Marcus always asks people to sit with for a moment.
What percentage of your new business comes from people who researched you before they called? What would it mean for your revenue if AI was actively steering those people toward a competitor instead of you?
We're not talking about a dip in web traffic. We're talking about being removed from consideration entirely at the moment a buyer is closest to a decision.
For most businesses, that's not a marketing problem. That's an existential one.
The importance question is the one that reframes everything else. You have a list of business priorities right now. New hires. Product development. A rebrand. A campaign. Where does "AI is filtering us out of buyer conversations" rank on that list?
Because if the answer isn't near the top, it's worth asking whether any of the other priorities matter as much as you think they do if the pipeline dries up because buyers can't find you when they ask an AI who to trust.
That's the context Marcus sets before he answers the how. And it's the context I want you to hold as you work through the five questions below.
Because the decision isn't really DIY versus DFY. It's: given what's at stake, what's the right way to get this done as well as possible, as fast as possible?
It depends on one thing: whether you'll actually do the work to Fix Your AI Trust Signals.
Between Marcus's experience on stage and the data we've collected, the biggest predictor of whether a company improves their AI authority isn't the size of their team or their technical sophistication. It's whether the right person owns the problem and has the time and conviction to see it through.
So answer these five questions honestly. Not optimistically. Honestly.
Question 1: Do you have someone in-house who has actually implemented JSON-LD schema before?
Not someone who has read about it. Not someone who is technically inclined and could probably figure it out. Someone who has done it, knows where it breaks, and understands why mismatched on-page content and schema code is worse than no schema at all.
If the answer is yes, schema-related signals are genuinely DIY-able for your team. If the answer is no, this is the single highest-risk area to attempt yourself. Done wrong, it doesn't return you to zero. It can actively signal distrust to AI models.
Question 2: Does your content team understand semantic structure for AI extraction, not just SEO-friendly writing?
There's a difference between writing that ranks and writing that gets cited by AI. Most content teams are trained for the former. Answer-focused semantic structure, the kind that lets an AI model extract a clean, confident answer from your page, requires a different set of instincts. Question-based headings. Self-contained answers. Conclusion-first paragraphs.
If your team already writes this way, restructuring your content is a realistic DIY project. If they don't, the output will read right to a human and wrong to a machine, and you won't know until your next score shows no movement.
Question 3: Can you dedicate 8 to 10 hours a month to this consistently, for 90 days, without it competing with revenue-generating work?
Trust signal improvement is not a sprint. Signal changes take 30 to 45 days to propagate and be detected by AI models. That means you need sustained effort across a full quarter before you see meaningful score movement. If your marketing team is already stretched, this work will be the first thing that gets deprioritized when a campaign goes sideways or a client escalates.
Be honest about your capacity. A half-executed improvement program produces results that look like failure, even when the strategy is right.
Question 4: If you improve your score and don't see movement in 60 days, do you have the internal capability to diagnose why?
This happens. Signal propagation has latency. Detection methodology has nuances. Some improvements show up immediately in a rescan; others take longer to reflect in how AI models actually respond. If your team can't distinguish between "we did it wrong" and "we did it right but it takes time," you're likely to abandon a working strategy too early or persist with a broken one too long.
Question 5: Is the cost of doing this slowly, or doing it wrong, lower than the cost of having an expert do it in half the time?
This is the question most people skip. They frame the decision as DIY versus spending money. The real frame is: what does delay cost you?
Every month your trust signals are failing, AI models are filtering your brand out of recommendation consideration in favor of competitors who have done the work. That's not a theoretical future scenario. It's happening right now for the majority of brands in every category we've analyzed.
What the honest answers tell you
If you answered yes to four or five of those questions, the DIY path is genuinely right for you. The AITS platform gives you the investigation reports, signal scores, and Score Accelerator playbooks to execute with confidence. Start with your free AI Authority Score, identify your highest-priority signal gaps, and work through the Technical Tier first. The platform is built to guide you.
If you answered no to two or more, that's not a failure of ambition. That's useful information. The brands that improve fastest aren't always the ones who do it themselves. They're the ones who are honest about where their time and expertise create leverage, and where they don't.
That's what our Done For You program is built for. It matches you with expert implementation partners from our agency network who have run our Score Accelerator playbooks before, know where the edge cases are, and can execute in a fraction of the time it would take a team learning the methodology from scratch.
This isn't outsourcing. It's deploying the right resource for the job. And given what's at stake, the question isn't whether you can afford it. It's whether you can afford to wait.
Where to go from here
If DIY is right for you: start with your AI Authority Score and the AI Trust Signals platform. Your signals will tell you exactly where to focus first, and MarcusAI gives you a detail action plan for the first 7 days and through the first 90 days of imnprovement, right in the tool, with your subscription.
If DFY is the honest answer: visit our Done For You page to learn how we match you with the right implementation partner from our agency network. Or if you're ready to talk now, book a call directly below.
Either way, the score is the starting point. You can't fix what you can't see.
Comments