For many business owners, the expected role of a consultant is to position a project and push it forward. After all, that’s how you get paid, right? Yet in the world of AI consulting, where you’d think an opportunity would be met with “strike while the iron is hot,” it’s unexpected to receive the recommendation to slow down, wait and not act (for now).
This is not the go-to playbook for consultants. There’s no fear-mongering about competitors getting ahead. There’s no timeline shave. Instead, there’s a simple explanation about how jumping in now would be wasting money and complicating things down the road. The rationales behind this suggestion provide insight into what makes AI projects successful, and what dooms them before they even start.
The Data Problem
For most companies, what’s most complicated about AI before it begins is the data. Owners think that after years in business, they have enough data to drive insights. They have customer lists, transaction histories, operational transaction notes, spreadsheets out the wazoo.
But AI doesn’t need data. It needs accessible, clean, usable data in relevant formats. A consultant doesn’t look and assume they’re good to go; they look at what’s available and see fragmented records across disparate non-communicating systems. They see Jane, Janet, and Janette as three separate customers; sales reports in one system while inventory reports live in another; relevant information hidden in PDFs, on sticky notes and handwritten memos.
Ultimately, one of the biggest reasons ai experts tells an organization to wait is due to data readiness. If the foundation isn’t ready, then whatever AI tool gets built on top will operate as garbage in/Garbage out.
Sometimes companies need three months to sort and standardize it. Other times they need a year or longer to operate systems integrations and process reconfigurations. If a consultant tells a company to start when they can just waive the data prep step, they’re essentially charging for implementation twice – once for attempt once it’s up and running to rectify the problems that should have been fixed during data prep.
The Team Problem
Another reason consultants recommend waiting is due to people. AI projects need people on them as dedicated time; owners expect too much from employees who already have full-time jobs and think it will only take a few hours here and there as assigned work. But someone must facilitate collaboration with the guidance of AI consulting, lend expertise, trail systems, provide feedback, and ultimately run the systems after implementation.
For most companies, this actual time commitment is drastically underestimated. While they think they can spare 1-5 hours a week during a quarterly budget meeting or holiday season transition, it’s more realistic that relevant persons will need 10-15 hours per week for 10+ weeks minimum, or longer, for buy-in.
Therefore, if a company is about to approach year-end closing, undergoing systems changes to improve production capabilities or drumming up resources for another initiative entirely, then bringing AI on board will be a perfect storm. Assigning it half-attention causes missed deadlines, poor communication, and inevitable stalls where everyone onboard feels defeated and convinces everyone else how worthless AI was as an effort.
Thus, a consultant may advise waiting until the fiscal year ends, until one key hire is made, or until the subsequent initiative is wrapped up. Not to say that it should never be done – more so it should be done when it can be done right.
When Leadership Isn’t on The Same Page
One of the hardest discussions for consultants to have is one with leadership buy-in because it’s not guaranteed for all parties involved. Maybe the CEO is excited about implementation while the CFO wants to keep costs down; maybe department heads are gunning for new positions and see the AI budget as one they could repurpose for their own needs.
AI projects that get launched without genuine support across leadership rarely make it past the first speed bump. When they encounter complications (and they will), the project gets deprioritized or shelved altogether. The sunk costs of time and investment become a cautionary tale and suddenly everyone who worked on it last time doesn’t want to touch it again.
A consultant who recognizes this dynamic early on will recommend more work internally before any technical decisions are made; these could include upper-management workshops to discuss benefits versus concerns, financial modeling and projections that are more realistic for needed buy-in or pilot programs that show reluctant executives what’s possible if they want to convince their peer leaders.
The recommendation not to move yet has nothing to do with lack of timing – but everything to do with ensuring it’s not set up politically, not technically, for failure from day one.
The Technology Problem
Sometimes it’s just too early for the appropriate use case. The technology isn’t there yet, or it’s there but too complicated or customized or expensive for a business their size to justify.
This occurs more often in niche markets or environments needing very specialized use cases. The reality is that these businesses either don’t have tech available yet commercially as off-the-shelf solutions or what does exist is less cost-effective than what would be standard in fields known or read about elsewhere.
The recommendation to wait comes with caution; wait until technology matures or stabilizes more; wait until the market fills itself with low-hanging fruit solutions; wait until costs come down substantially. It’s not so much indefinite delay – more strategic timing based on trends outside of anyone’s control.
What Waiting Actually Looks Like
So, when consultants recommend waiting, they don’t mean doing nothing in the meantime. They suggest getting ready for what’s next. This includes cleansing data better prepared for collaboration during implementation; documenting processes better standardized for implementation efficiency; training staff in basic AI methods so that everyone speaks the same language; even small sub-projects that don’t require major outlays yet still provide solid insights.
Therefore, when implementation does come down the pike, it goes much quicker since complications have already been avoided through prior phases. Everyone knows what’s expected of them; data is already good to go; there’s leadership support and timing makes sense from an operational perspective.
Ultimately, it’s counterintuitive that this quicker implementation occurs down the line instead of immediately but what happens when rushed causes backtracking that never allows firms to move faster than if they’d only waited a bit during prep stage beforehand.
The Decision Making
Not every “wait” suggestion is justifiable by any means, some consultants play it too safe; some companies have circumstances emerging that make an immediate start worth risking avoidance plans. Therefore, those getting pressured should evaluate specifics before using these costs against them.
Market realities are important, if companies are missing out on market advantages for every step an AI process takes then waiting could annihilate everything they’ve built up over time thus far. If companies are dealing with immediate operational challenges where they can implement what can be resolved via AI now makes more sense than anything else, then urgency trumps waiting.
However, for many companies who are inexperienced implementing AI projects at all stages, the recommendation not just to “wait” but to get prepared also allows for these uncertain elements of costs involved which otherwise would’ve ended up as lessons learned through taught experiences otherwise. It’s important not to just implement AI for good production quality but AI that’s actually usable for real value going forward. Sometimes that means taking two steps back before pushing forward.
Read Dive is a leading technology blog focusing on different domains like Blockchain, AI, Chatbot, Fintech, Health Tech, Software Development and Testing. For guest blogging, please feel free to contact at readdive@gmail.com.
