top of page

Finding an AI Partner is way too Complicated

There is usually only one question that crosses a business’s mind when considering AI: How?

The answer, unfortunately, isn’t as simple.

See, artificial intelligence alone is complicated — there are many subsections that fall under the umbrella of AI, and infinite ways AI is applied to everyday issues.

Just wanting to adopt AI isn’t enough, though. The steps from here to full adoption are varied and tedious, but that doesn’t mean finding someone to help you should be.

AI’s biggest problem is its reputation

AI doesn’t exactly have the greatest reputation. Between controversies surrounding deepfakes and concerns of personal data privacy, and the theories begin. Adding an element of manmade humanity sparks misconceptions that spread like wildfire.

Luckily, we were able to sit down with our very own CTO, Garry Chan, to discuss the issues AI as an industry currently faces and the single solution that could finally get AI startups and the potential adopting companies to align.

Why finding a partner in AI is tough

Companies don’t understand AI

Most businesses don’t know enough about how AI works to even begin making educated decisions on where to start, such as the partners they need and what to look out for in said partners. As usual, the miscommunications usually come from the overuse of technical jargon from AI startups.

“The AI solutions provider has to explain and quantitatively prove why AI can help serve the customer's customers better and faster, through the richness of data and deep tech solutions,” Chan advises. “Go back to the basics and think about… what problems keep them up at night.”

Companies don’t see immediate benefits

The problem runs deeper for AI startups who have a hard time explaining the benefits of AI to prospects — because they have difficulty proving the golden metric, too: ROI.

“I believe customers always know best,” Chan mentions, “whether or not you're selling AI-powered solutions,” he adds. “Personally, I would focus on these themes [when selling to prospects]: quantifiable pain points, your unique value proposition, and ensuring your communications stay solution-focused.”

A deal made isn’t a deal done, though. Even after AI is adopted, it can take upwards of 18 months to see real results. Add on that the onboarding process is a tedious, capital-consuming process, and it’s natural for any company to be wary. AI is an investment for sure, but it’s not a short-term solution by any means.

Companies don’t trust AI — or the companies selling it

We touched on this earlier, but it remains true for far more than just individuals. A little more than half of leaders worldwide trust AI to make decisions… most of the time. And while 50% might seem good, a divide like that in any company is not.

“If a customer does not have AI today, think about the array of options available [to solve their current problem],” Chan said. “e.g. [do they] keep things the way they are, buy expensive enterprise solutions, build a team of AI experts themselves, outsource the work, etc. Then think about where the AI solution ranks among these alternatives.

The market is oversaturated

Here’s an embarrassing fact: Roughly 60% of “AI startups” don’t actually use AI in their businesses.

With all these new startups sprouting in the market just to be on trend (or so it seems), it’s naturally difficult to find a single AI partner worth investing in. Many companies end up talking to many specialists rather than one entity.

The AI provider has to act as the customer's AI product manager, in this case, identifying and prioritizing AI use cases based on feasibility of the solution, value proposition, time to market, cost of implementation, etc,” Chan says.

Companies think they’re “too small” for AI

“They may feel AI is only for organizations with big budgets, well-structured data, and a deep bench of in-house talent ready to engage in data science and AI/ML activities,” Chan says. “I think the AI provider, even if it is mainly product-focused, has to take a holistic, solution-based approach to help address the customer's pain points.

For example, … start with an assessment of the organization's AI readiness first, such as the availability of data, talent, and processes. Next, define a roadmap for onboarding and engaging the customer, its users, and its stakeholders. Once the solution has been deployed, define a process and relevant instrumentation for ongoing monitoring and continuous improvement.”

The solution? An age-old business skill

Long-term thinking. Strategy. Intent. The ability to see past a list of tactics and into the greater overall picture is the idealized (and majorly romanticized) role of a visionary CEO.

Putting away images of starry-eyed dreamers, though, and long-term thinking becomes the simple solution to one of our most complex emerging problems in business. And when paired with a competent AI partner, long-term thinking can greatly reduce the onboarding process.

“Here’s why:

  1. Organizations want to focus on what they are good at, and will/should consider offloading the remaining tasks.

  2. Organizations can accelerate their learning and adopt best practices by working with qualified partners.

  3. Organizations can mitigate the risks of solution development and deployment by bringing on qualified third parties rather than attempting to complete every task in-house.”

Just like hiring your first employee is a terrifying investment, onboarding AI is certainly a scary investment — at first. The beauty of the experience is that the more you do it, the better you get.

Your success in AI depends on how big you can think

“In my view, in order for AI to be successful, we need to tap into the wealth of enterprise-level, cross-functional data that an enterprise has accumulated about its customers and operations over time,” Garry Chan suggests. “While AI can be deployed in a focused use case e.g. chatbot, it can be much more powerful and useful if we adopt a broader view.

[For example,] if an online retailer could consolidate the data collected from its’ e-commerce site, CRM, financial systems, and support desk and apply AI judiciously, it could create a more comprehensive profile of its customers for future marketing, customer engagement, and support initiatives. The strategic view naturally requires longer-term thinking and implementation roadmap.”

Finding a partner in AI may not be so tough after all

Unsurprisingly, it seems the answer comes from both sides of the partnership: Companies interested in adopting AI need to be ready for a time- and capital-intensive process, with big dreams to carry them through the grueling R&D process. And AI startups could get much better at explaining what they do in simple, straightforward terms.

Obviously, a company can always adopt AI without an AI partner,” Chan mentions, “However, for a whole host of reasons, such as risk mitigation, best practices, time-to-market, and access to talent, I feel that the use of an AI partner is not an ‘if’ but a ‘when.’”

Luckily for us, companies like AI Partnerships Corp. exist. With over 120 partners, deciding to work with AI is deciding to be immediately plugged into an ecosystem of 120+ pre-vetted technology and solutions partners.

Next steps for adopting AI starting today

As an added bonus, we asked Garry what he would focus on if he were investing in AI from the very beginning. Here’s his list:

  1. Team. Define the internal stakeholders and ‘owners’ that will sponsor the initiative, define the goals and objectives, and project-manage the process. Based on strategy, decide how you’ll move forward:

    1. build in-house

    2. use external partner

    3. or a hybrid model

  2. Customers. Define who the ‘internal’ customers of the AI solution are, their pain points, and their acceptance criteria.

  3. Product. Define the ‘what’ of AI — identify candidate use cases, value proposition, costs, timeline, and KPIs.

  4. Budget. Define the total costs of the AI solution, whether it's done in-house or bought.”

60 views0 comments


bottom of page