Pricing for complex or project-based projects and services firms is challenging, involving vast reams of data, intricate calculations, understanding of market dynamics in low-volume markets transaction-wise, and customizing the pricing to match your client’s requirements. Pricing for Government contracts adds more complexity in terms of justifying the cost (called the ‘basis of estimate’) and avoiding fines for false or untrue costing. Did you know the top aerospace & defense firms have been collectively fined well over $1B in the last three years for pricing mistakes?
Traditional methods of pricing for large enterprises in industries selling projects such as professional services, aerospace & defense, hi-tech, construction and complex manufacturing rely too much on intuition, tribal knowledge and Excel spreadsheets. Professional services firms, in particular, have for decades relied on the classic ‘time and materials’ based pricing model where billing rates agreed with the client are multiplied by estimated hours to complete the statement of work to derive a total price. As more and more firms are selling combined offerings including engineering, manufacturing, hardware, software, spares, maintenance and support, and as your clients are becoming more adept at transferring risk down to you, their vendors, by demanding fixed price or outcome/value based pricing, the challenges and opportunities to get pricing right for project and services-based firms have both increased exponentially in the last few years.
Artificial Intelligence (AI) offers a data-driven, systematic approach to optimize pricing decisions. AI can more efficiently sift through vast amounts of structured and unstructured data in both the client’s RFP request documentation, and in your rich prior performance and cost/price history - which is typically locked up in your enterprise resource planning (ERP) system. AI can apply machine learning and predictive analytics techniques to your cost and price history, as well as large language models (LLMs) to significantly speed up your response text and answers to the requirements and questions in your client’s RFP.
In summary, leveraging AI for project and services-based pricing will bring the following benefits:
Dynamic Pricing: AI based dynamic pricing algorithms can analyze value-based factors, client or market-facing factors, competition and of course your historical pricing for the same/similar products or services in the past to ensure your pricing remains competitive while maximizing profitability.
Accurate & Confidence Costing: AI based costing algorithms can replace tribal knowledge (such as “I remember Fred did a similar project to this four years ago, let’s talk to him about the cost”) with advanced search and re-use, while using predictive analytics and other statistical methods to understand cost drivers and sizing inputs for your new proposal.
Faster & Better Proposals: AI based proposal response text generation, and answering basic questions in the RFP such as “Do you have ISO27001 compliance?” can significantly cut down the time spent by your proposal team trawling through company content and aligning it to the RFP. This allows your team to spend more time on adding value and personalizing the response to the client/RFP requirements.
Enhanced Profitability: do you think AI could enhance your pricing, on average, by 1%? Or put another way, do you think that you are pricing one tenth of your contracts at 10% below what you could price them at, and still win? For a $4B revenue project/services firm that’s a $40M annual return to the bottom line with minimal cost.
Customer & Competitor Segmentation: every client and RFP looks unique, but in reality you are always selling something very similar to something you’ve delivered before, else why bother? AI can segment your customers and markets in terms of willingness to pay, and better analyze your competition to understand where you need to sit price-wise vs. your competition, resulting in client-specific tailored pricing strategies which maximize value and profit margin at the same time.
Business Planning & Forecasting: most large firms have sales and finance operating in separate silos and only consider the pipeline or submitted proposals vaguely in business planning. It is an inexact science, but AI can simulate pricing scenarios, consider trends, risks, and supply chain or talent acquisition/availability constraints, to optimize a data driven revenue, project-cost and resource-needs forecast for the entire business.
At the same time there are challenges in adopting AI tools and techniques in pricing, such as:
Data complexity and quality: most of your master service agreements are in emailed PDFs on someone’s laptop, right? Even system data such as historical timesheets in your ERP might not be that useful (who books time to the lowest level task?).
Complex projects or services offered by your firm with lots of options, parameters and variables as opposed to cookie-cutter repetitive production of standard widgets.
Legacy systems such as ERP systems which can be hard to get data out of and may not provide easy access to performance history data.
Regulatory and ethical concerns - AI driven pricing needs to be transparent, fair and ensure compliance with relevant regulations such as DFARs for aerospace & defense firms.
User buy-in: pricing especially for partners in professional services firms is intricately linked to the partner’s revenue and client satisfaction. Not only do these people need to trust AI recommendations, but in some cases will need to be convinced to share their data with it in the first place.
With this in mind how do you start the journey of implementing AI-driven pricing?
The first piece of advice is to have a strong bias for action. Even if you do entirely the wrong thing initially, the experience and learning will be invaluable and far better than spending a year waiting or considering options, meanwhile your competition implements AI-driven pricing and leaves you in the dust. Beyond that, we recommend the following:
Engage with or hire an (outside) expert. You can learn a lot by using ChatGPT (yes, it helped with this blog!) but it’s always good to supplement this with an expert. Whether this expert is a consultant or employee who has spent twelve months researching and implementing AI-projects is up to you, but there’s nothing like getting expert help
Start small, then scale slowly. You will make mistakes and rework stuff, especially in a relatively new domain like AI-pricing. Select a subset of your products or services to implement AI-driven pricing strategies, or design a strategy which is initially based on prior pricing history alone and factor in market value, client’s willingness to pay and competitive landscape in a later phase. Per point no.1 the lessons learned will be invaluable
Adopt a data and user-centric approach. The old adage “garbage in, garbage out” cannot be truer than for AI, because AI is awfully efficient at automating bad processes or drawing the wrong conclusions from inaccurate data. Invest time and resources to clean-up your data, and put in place data quality and governance processes across your organization, not just for AI. At the same time, consider the process from your user’s perspective, how easy will it be to use any new tools and what prerequisites are needed?
Don’t remove the humans. AI won’t do it all for you, there is no ‘easy’ button as it were. You still need to hire pricing experts who have some good ideas on how pricing should be optimized without AI tools. Then you will not only understand what you are trying to achieve, but you can measure success as you go down the right path.
Invest time and money in the right infrastructure. There are several LLMs (large language models) and data-science tools you can use, and several companies provide specialist tools, databases and hosting for AI-driven technologies. In general security and privacy concerns are quickly diminishing (for paid account users) but be cognizant of the potential need for in-house hosting if you deal with sensitive or international trade restricted data. Cloud-based systems are recommended, while at the same time simpler and smaller LLM’s can now be hosted in-house for a reasonable cost.
Investigate AI-based CPQ tools. Software vendors for pricing tools such as CPQ are typically on the leading edge of new technologies, and AI is no exception. Whether you look at PriceFX for product-based pricing, or tools like Twenty5 for project/services pricing, it pays to talk to these vendors, get demos and consider acquiring their tools. Not only will you find these vendors provide free outside expertise/ideas for you, but they have often solved some of the thornier challenges like integration to legacy systems such as your CRM or ERP, data clean-up and training data, regulatory compliance and even user buy-in via well thought out user experiences and change methodologies.
Follow Twenty5 on LinkedIn for more updates, content & news!
コメント