AI – along with machine learning and automation – has been a part of the enterprise IT lexicon for some time. However, recent innovations and tailwinds from the pandemic have accelerated their adoption. That’s led to greater focus on cost control. Many companies are looking for ways to keep AI pricing and costs in check as they embark on new AI-related projects and expand those that are delivering value.
Historically, many of the AI purchases NPI has analyzed have been related to internal analytics and BI tools. More recently, we’ve seen a noticeable uptick in customer-facing systems including chat bots, augmented/customized CX, and AI-assisted search/insight engines (vendors include eGain, Qualtrics, and Amelia). The market and associated enterprise spend is among the fastest growing IT segments. Analyst firms tracking market dynamics pin the total market for AI solutions to be above $300B in 2021, with 20 percent CAGR expected for the next decade. AI software comprises much of this, but the currently limited market for AI-embedded hardware products is expected to grow at an even faster pace.
The solutions and capabilities that fall into the AI category are evolving at warp speed – and are quickly becoming ingrained into the technologies that enterprises use. While many of these capabilities still have a way to go before reaching “true AI” (or “strong AI”) potential, the automated assistance that newer tech can bring to workforces is no longer a novelty. In fact, it’s increasingly becoming a necessity.
This also means that vendor sales teams are eager to maximize revenue during this phase of market expansion. The result is AI pricing and cost control have become challenging waters to navigate. Below are suggestions and considerations to help enterprise IT buyers keep spend in check.
Spot Check Key Metrics that Impact AI Pricing
To be cost-effective, AI-related solutions must be monitored and adjusted, particularly after an initial deployment term. NPI recommends checking a few common spots to validate utilization rates in AI-specific areas:
- Attended vs .Unattended Bot Usage – Two applications of Robotic Process Automation (RPA) are Attended and Unattended Bots. Unattended Bots usually cost a fraction of the fees for Attended Bots. NPI recommends performing price benchmark analysis on both types to ensure pricing and price ratios are within fair market value parameters. (Typical vendors NPI analyzes for clients include UiPath, Automation Anywhere and Blue Prism.)
- Efficacy Over Time – Systems should improve engagement and produce noticeable increases in KPIs that are defined up front. The vendor should communicate continuous improvement, especially in larger solutions with 6 or 7-figure (or more) price tags.
- Successful Customer Support Mitigations – For customer support-oriented AI solutions such as WalkMe and LivePerson, one metric is how many successful mitigations the solution was/is able to perform. Chat bots or responsive AI-generated menu choices are usually measured by their ability to reduce reliance on human customer support representatives, and these statistics should be checked ahead of any major renewal or qualified ahead of first-time purchases.
Don’t Rush into Long Term Commitments Without Due Diligence or POC/Pilots
Despite AI-centric solutions being limited in scope, recent growth in the market has many sales teams pushing for long-term commitments from the go – usually with some sort of pricing incentive tied to the extended agreements. However, NPI cautions that the industry is likely to go through more consolidation and advancement in the coming years, meaning extended agreements are likely to benefit the vendors more than the customers.
If a long-term agreement is the preferred outcome, NPI recommends (at a minimum) making sure the incentives included are in line with appropriate concessions. We also highly recommend consideration of a shorter starting term with a formal POC/pilot to validate use cases.
Establish AI Pricing to Scale Appropriately, Check Product Roadmaps
Part of the value of a pilot is a slow approach, but you don’t want to lock yourself into an uncompetitive unit price as solutions scale. The scalability of modern AI solutions is one of their major benefits, a nice change from the days when lack of scalable features and inflexible AI pricing were a barrier. NPI recommends getting a sense of what kind of volume discounting could be had in a “best-case scenario” where adoption of a given solution could hit the highest realistically expected levels.
Product roadmaps are an important item to review with vendors during initial purchases as well as renewals, particularly in cases where vendor consolidation has occurred. NPI suggests checking to see how any planned changes in future products could impact current spends, as AI-related deals can see a greater impact from changes than other software spends, especially if any main metrics change.
As the competitive landscape evolves, AI pricing dynamics are evolving too. Staying on top of changes and considerations – as well as validating you’re paying at or better than market for AI solutions – is crucial as customers begin to form a solid AI footprint within their business.
- Bulletin: 10 Reasons Why IT Buyers Should Perform Price Benchmark Analysis
- Blog: RPA License Costs: Keeping a Lid on Your Spend
- Blog: Combatting Objections During IT Vendor Negotiations
- NPI Service: IT Price Benchmark Analysis & Contract Negotiation Intel
- NPI Service: SaaS License Optimization Assessment Services