How AI Influencer Overreliance on ChatGPT Hurts Marketers
The Swiss Army Knife of AI has many uses, but it lacks the domain focus marketers need to succeed.
AI influencers come in many flavors. Most are solopreneurs or influencers who build a brand by creating content, sharing opinions, and tips about AI. And, most use ChatGPT as their primary LLM and share their prompting tips and insights on how it helps them and their clients' marketing. That makes sense as they cannot afford or access enterprise-grade AI solutions.
This overreliance on ChatGPT is likely hurting marketers, who would benefit more from domain-specific tools and mastering the limited tools they have access to through their existing software. Don’t get me wrong, ChatGPT is the top LLM for consumer use (2.5 billion prompts a day!). But what is good at home is not necessarily good enough for work.
Business-Specific Reasons to Question ChatGPT Dominant AI
ChatGPT is just one tool, a singular dot, in the MarTech Map. By the way, martech founder Scott Brinker is our next guest on the No Brainer AI podcast.
While ChatGPT can perform many tasks, it is often inferior to tools developed specifically for marketing tasks (domain-specific). It is essentially the Swiss Army Knife of AI.
Instead, an organized marketing department would best benefit from identifying the AI tools tuned for the specific job. Here are some examples:
Consider MarketMuse or even your native CDP’s AI capabilities for customer personalization.
Customer service trends can be better analyzed with Alteryx.
Data visualization is better in Flourish or DataWrapper.
On the content writing side, consider how far domain-specific LLMs Jasper and Writer have come.
From a visual content standpoint, you can use Canva AI Magic or Adobe.
When marketers have control over their own tool selection, they can access best-of-breed solutions for their specific tasks (as opposed to a tool that performs most tasks on an average or below-average basis). These AI apps should be considered disposable. Brands can always implement new solutions as they better execute the need. And with AI continuously evolving, it seems like there is a better tool almost every quarter.
There is another class of marketers whose enterprises simply will not provide access to ChatGPT for a variety of reasons, ranging from valid security concerns to the high cost of enterprise-grade licenses. They are left to use Microsoft Copilot for native Office software, Adobe’s creative AI tools (which are quite good, as noted), native AI in HubSpot, and so on.
In these cases, the best advice and tips marketers can get will be about maximizing existing tools and their AI capabilities. But instead, they are told how great ChatGPT is.
Simple Use Cases Yield Underwhelming Results
Most of these influencers fail to measure AI outcomes in anything other than time saved or productivity increases. AI influencers tend to promote productivity increases with ChatGPT, including time saved, enhanced productivity, and increased speed. These tactical measurements are KPIs, but not accurate barometers of marketing strategy success.
Certainly, these are useful for executing daily tasks on the front line, and are usually the easiest and fastest metrics to measure. They also illustrate ChatGPT’s primary use case limitations as a productivity AI tool for individual contributors.
This has spawned quite a bit of debate recently, best typified by the MIT study that found 95% of all AI pilots fail. The MIT report’s unspoken part is it only focuses on delivering hard ROI within six months, and according to only 50+ CEOs. But the larger conversation about outcomes brings us back to use cases and measurement.
Operations is the lowest-hanging fruit, and the right way to start and show initial AI successes. "Yes, AI can work." However, sales and marketing is the lifeblood of any company. No business exists without sales, and failing to address the ROI question is, in my opinion, a form of malpractice.
Marketing should always seek to create a tie to ROI or show how the action helps the organization achieve more of it (customer acquisition cost or CAC). Enterprise AI marketing use cases illustrate that stronger marketing-oriented outcomes can be achieved through content personalization, funnel optimization, optimized advertising, customer insights and segmentation, and customer-facing AI solutions. You don’t see AI influencers talk about these types of use cases very much, instead preferring to highlight their 18-page branding prompts or the like.
Even with productivity, speed to market, and increased quality use cases, a marketing department should yield more business, better customer relationships, and new leads. Do the extra work and show a tie.
Sophisticated marketing AI use cases cannot be easily built for brands using an LLM in their web-facing product interfaces, though LLMs can be integrated as part of more complex AI models or as private instances to achieve marketing use cases.
Better sales and marketing use cases for AI are harder and require discipline. They require data governance, strong processes, and strategic thinking. And, marketing executives need to lead their organizations through the necessary transformations, rather than simply purchasing technology.
AI Productivity Theater
Look, AI influencers have some value. They have demonstrated how ChatGPT (and, by extension, LLMs) can be utilized by individuals and small teams in operational and marketing workflows.
Their constant focus on how to use ChatGPT makes for great productivity theater. It’s certainly useful toward building the perception of expertise. However, the guidance is tactical and limited in scope. This is particularly true when guiding marketers on how to use AI.
Instead, begin with this baseline: Every marketing AI conversation should be guided by your mission. In most cases, this is driving ROI through brand building and go-to-market actions, as dictated by your overarching corporate sales strategy. With this stance, tools become disposable (as noted above) and strategy dictates which use cases and tools are selected.
Beyond the above-listed reasons, there are several reasons to be concerned with ChatGPT-dominant guidance:
OpenAI may not have the best tech anymore as it is losing the B2B developer and enterprise markets to Anthropic. Think about that. More and more software applications are actually using Sonnet or Opus instead of GPT. If developers and large enterprises think Anthropic is superior, are marketers getting the best LLM guidance from ChatGPT-centric consultants? If Google Gemini has the best Deep Research, why use ChatGPT?
Security concerns with OpenAI are very real. They are a primary driver for developers and enterprises procuring other solutions like Copilot and Anthropic’s Claude. Even with guarantees from OpenAI, we have seen numerous breeches of trust and data security. Sharing proprietary data on the Internet is less traumatic for an AI influencer than a business with a CRM full of customers.
Conflicts of interest may be driving ChatGPT (or any other tool) suggestions. I know of one very prominent AI influencer in the marketing space who maintains a seamless relationship with OpenAI. This is not openly disclosed and creates a conflict of interest. Look, I get it - we all have motives for creating content. However, when you provide tool advice without disclosing relationships, it becomes very problematic.
Bias: What’s the motive here? Is it to help people adapt AI (as stated), or to build influence and get speaking gigs, etc. I know for CognitivePath, we create content as part of our own marketing efforts. It should be obvious to anyone on LinkedIn or any other online business conversation that people are publishing content to achieve an outcome. Even if this content is karmic and helpful, there’s a motive. Unpacking the motive helps digest the underlying bias implicit in the content.
I am sure there are more reasons to recalibrate AI tool advice from influencers. But let’s focus on the ultimate goal, becoming better marketers.
Tool Selection Guided by Strategy
Marketers are best served by revisiting their strategy before rushing to the credit card to buy ChatGPT because it’s the cool toy on LinkedIn. This is shiny object syndrome at its best, and while 18-page prompts sure seem impressive (or painful, depending on your perspective), this is not a good barometer for choosing an AI solution.
Select tools based on your specific marketing needs, short and long-term, and stay true to your North Star mission, building brand and sales opportunities. Measure accordingly.
By the way, ChatGPT might actually be the right tool for you. We use several LLMs at CognitivePath, including ChatGPT, which we used to deploy our AI Maturity Model GPT so companies can self-assess their marketing readiness.
How many marketing teams are making tool decisions based on influencer hype rather than actual business needs? And how many are overlooking powerful AI capabilities already built into their existing software?
A marketing stack audit is the best place to start, with the real guiding question being: what problems are you actually trying to solve? What are the right use cases to address them? Then work backwards to the tools — whether that's ChatGPT, your CRM's native AI, or something you haven't considered yet.
It's the due diligence process every marketing leader should run before adding another monthly subscription. More on this in the next article.