The panacea has arrived! With artificial intelligence (AI) my company can reduce costs, improve customer service and become more productive.
Or can it?
That is the prevalent storyline: AI will solve all your problems by providing access to key information faster and in context of the situation or conversation, which enables your sales and marketing teams to become more efficient, productive and engaging.
Nothing to it.
However, according to The National Observer, actual experience does not always realize this pie in the sky picture of AI.
- About 70% of managers say their direct reports have made AI-related mistakes this past year (Resume.org).
- 58% of managers said their direct reports submitted work that contained factual inaccuracies generated by AI tools (Resume.org).
- 25% of managers said they ended up missing deadlines as a result and 22% shared that these errors damaged client relationships (Resume.org).
With over 15,000 AI tools—some analysts estimating the number closer to 50,000—available today and growing, there comes plenty of opportunity to put this tool to work for your teams. And while there are efficiencies and greater productivity to be had, using AI effectively takes a specific approach and a diligent leadership team laying out a core strategy and structure.
Here, we share two of the biggest areas of pitfalls and opportunities for your organization to get ahead of today to make the most of AI in your revenue organization.
AI QUALITY MATTERS IN BOTH INPUTS AND OUTPUTS
Content, data accuracy and form consistency are more important now than ever before. Like any form of automation, AI tools are only as good as the data sources they access.
For example, an acquaintance was working within their sales operation team to identify key attributes of companies within their target market — things like revenue, earnings, number of employees and number of locations. They created a model with the output formatted the way they wanted and set the engine to perform general internet searches on the first 10 companies. They took the time to validate the results and were satisfied. Then they targeted the next larger tranche of companies and found that the data became inaccurate as the list grew longer. This is called “AI Drift.”
THE ISSUE: AI DRIFT
AI Drift refers to the degradation of AI models over time due to changes in data patterns or relationships. Drift occurs when there is a mismatch between data used to train a model and data encountered in actual production. In this example, the model found a source that was excellent for some company types but not as sound for others, and the resultant data was less accurate depending on the industry of the company researched.
THE FIX
To ensure accurate outputs, effective marketing and sales teams are focusing queries first on known internal data sources. To speed AI response times, they are also focused on formatting information sources between Word documents, PDFs and presentations in consistent ways.
For example, product documentation can be created in a specific, repeatable format. Client value stories can follow a similar approach. Value messaging frameworks on personas and value drivers can follow specific patterns. Value propositions can follow a leading-practice formula. The result is that human consumers are better able to digest information in enablement sessions or as external website users, and AI tools become “trained” to more rapidly access data and respond to queries.
This is well within reach. One client of ours successfully directed their AI sales tools to their library of product marketing materials. The first step was ensuring these materials all followed the same leading-practice format. They eliminated the need for numerous time-consuming searches by the sales team and replaced that with context-based queries that could be used in advance of and during customer / prospect engagements. This resulted in richer and more robust discussions that were based on the latest and most accurate information. And for the marketing team they had a real return on their investment of time to create the assets. They had found a way to break the issue of the sales team not being able to locate the information they need when they need it and how they need it. Context and access create value.
INFORMATION TECHNOLOGY MUST PLAY A ROLE IN YOUR AI APPROACH
Not dissimilar to the advent of Software as a Service (SaaS) offerings, AI tools are sprouting up across the organization and across users rapidly. With SaaS, companies found themselves acquiring software solutions in a distributed fashion by department or function with little regard for standard IT decision-making criteria like ease of integration, maintenance processes and upkeep, operational security requirements and support requirements.
Over time organizations have trended towards a mixed model of decision-making with IT performing key evaluation functions to identify an approved set of vendors or solutions, while the business unit made the final tool decision. A similar approach must take place when it comes to AI versus a rogue-like approach that hinders value realization.
THE ISSUE: AN INFLUX OF TOOLS
AI tools can be desktop- / laptop-based, internet-based OR acquired as part of more standard SaaS solutions. The rapid proliferation of tools creates numerous risks from the mundane to extremely concerning.
On the more mundane side, there are concerns with cost, reliability, maintenance and accessibility. For instance, many tools reside on the user’s laptop, and the models are built and reside there as well. What does this mean for your organization when an employee is sick or on holiday? What does this mean if they leave the job or company? How will new employees access and learn about these tools? Who will fix script errors or maintain them if data sources change?
On the more extreme side, tools could expose sensitive and confidential information or worse create cyber security risks with the environments they access or the software they deploy when downloaded. Therefore, AI can create crippling liability issues.
THE FIX
Banning BYOAI (bring your own AI) can cause more issues than it solves if users do not have good tool alternatives to choose from but are given ever-increasing productivity pressures. The best approach is to team up with IT (a la SaaS) on procurement of AI solutions and effectively enable and educate employees.
IT needs to determine the best tools and/or acquire tools that meet data and operational security requirements, and they need to have a keen awareness of the user needs and applications to promote adoption of tools versus BYOAI. A high-quality AI tool demonstrates transparency, accuracy, stability and strong data protection measures. The most trustworthy platforms frequently update their models, publish clear usage documentation and maintain responsive support systems.
Creating a “menu” of AI tools that are safe to use can reduce workarounds and their resulting risk and costs. Training in AI usage best practices can accelerate user adoption and help educate the organization to security and brand risks. Do not assume that age is an indicator of AI competency. Instead create a best practice approach to getting the most value, and offer formal training and guidance. This approach also enables users to share best practices and approaches in a supported environment minimizing the risk of turnover disruption created by BYOAI.
THE AI STORY OF SUCCESS CAN BE REALIZED — WITH YOUR LEADERSHIPS’ DILIGENCE AND BEST PRACTICE
The organizations seeing real gains from AI — efficiency, cycle time reduction, increased innovation, better decision-making and more — are not chasing every new tool or delegating responsibility to technology alone. They are investing in content discipline, data quality, thoughtful tool selection and cross-functional partnership — particularly between business leaders and IT. They are training their teams not just on how to use AI, but when and why to use it in the context of real client interactions.
AI’s promise is real. But realizing it requires leadership that is willing to invest and put the right foundations in place. Done well, AI helps sales and marketing teams become more responsive, more relevant, and more effective. Done poorly, it simply helps them get the wrong answers faster.
The difference is not the technology. It is the rigor, governance and intent behind how you deploy it.
Contact our experts for an AI content readiness assessment.