Developments in artificial intelligence (AI) offer incredible opportunities for e-commerce merchants to work more efficiently and innovate. However, successfully integrating AI in e-commerce also brings risks that need careful management. Change needs to be planned and carefully executed to bring along all those involved. Here are some of the issues that your business needs to consider for a successful AI introduction.
Return on Investment
AI should be treated like other areas of investment in your business. Set explicit objectives because without them, it is impossible to know where you are going. To quote Alice in Wonderland when talking with the Cheshire Cat.
“Would you tell me, please, which way I ought to go from here?” “That depends a good deal on where you want to get to,” said the Cat. “I don’t much care where–” said Alice. “Then it doesn’t matter which way you go,” said the Cat.
Studies about the effectiveness of AI used in organisations over the last two years show consistent efficiency gains. Where a structured programme for introducing AI was used, a 10% to 15% increase and beyond in efficiency was seen. Where AI was introduced without training, efficiency gains have still been positive, but very modest.
A simple way to calculate a target ROI Is:
((Staff Costs) * (Efficiency Target: 5%, 10%, 15%)) – ((AI License Costs) + (External Consultancy Support and Training))
As an example, for 10 staff, with a cost of £400,000. AI licenses are £3,000 annually. External support from a consultancy is £10,000.
For a target of 10% efficiency gain.
((400,000)*0.1) – ((3,000)+(10,000)) = £27,000 efficiency saving
Policy
Your policies are the operating system of your business. They set expectations, offer guidance and outline what staff can and cannot do. They exist to manage risk for the business. These apply as much to AI, as they do to expenses or entertaining.
Creating an AI policy puts in place guard rails for staff. For example, it is highly likely your staff are using free AI from suppliers like ChatGPT. This creates risk for the business, as the terms and conditions state that any data entered is added to the wider AI data set. Confidential or sensitive information can unwittingly be made public.
You will also need to include guidance on the process to ask for different AI services.
Audit
Where to start? Any unofficial use of AI is a good place. The problems that your colleagues are solving represent pain points. They show tasks and processes that will benefit from using AI.
You should also know what AI offer is within the main tools used to run your business, such as customer support desk or CRM. If your tool providers don’t have plans for built-in AI, it might be time to consider whether they’re the still right for you.
Plan
Create short, medium and long-term plans. You will be revising the medium and long-term plans on a regular basis, but without them, it will be difficult to understand your direction of travel.
Be Experimental
AI is new for almost everyone; we are all breaking our own path. We can best do it methodically, planning small pieces of work. Start by setting clear objectives about what you want to achieve. Design an activity that uses AI. Measure what you do, success or failure. Make sure that the experience is fed back into the business.
Tools
While there is a large number of AI tools in the marketplace, for most organisations, the choice is straightforward.
If you run Microsoft systems, you will want to use Microsoft Copilot. If you have Google software in your business, Google Gemini will be for you. You are not just selecting the best technology, as this is context sensitive and somewhat subjective. You need to consider other issues; security, integration with existing systems and set up time. For example, adding Copilot to your Microsoft estate means that it inherits the permissions and access rights that your IT department already have in place.
Training
Essential for the successful introduction of any new technology, but especially so for AI. Create an AI driving license for your staff; a test of how well they know the AI policy and how to do common tasks using AI. Unless you have your own training department, consider engaging an external consultancy that can create custom training for you.
Develop an AI Playbook
You have a plan, trained staff and are starting to gain useful experience with AI. You need to capture this by developing an AI Playbook. It can be document but is more likely a SharePoint instance or a wiki or some other repository that users can add to and reference. It is a space to collect useful information; how to prompt efficiently, a prompt library for common tasks, references to useful articles and HOWTO guides. The playbook should be dynamic, with new experiences and techniques being added often.
Keep Learning
Once you have started to use AI, new applications and ideas will show themselves as you become familiar with the technology. You have started a pathway of continuous learning, to get the most out of your investment.
Encourage your staff to share their experiences and develop a community of practice within the business. There are many mechanisms for doing this, including expanding the playbook into an internal online community where people can ask and answer questions, or maybe a regular lunch and learn session each week. You will know what suits your business best.
There are so many ideas and so much new knowledge out in the world about how to use AI that even a casual interest in continuing to learn will show benefits. Enabling people to do a better job through access to AI, training and with a mandate to experiment, will reap rewards for you.
Navigating AI in e-commerce integration can be complex. Contact us today for a personalised consultation to discuss how your business can strategically adopt AI for maximum efficiency and innovation.