Differentiated pricing for Gen AI Consultants - Part 2
Deconstructing the Gainsharing Model for AI Consultants
Implementing gainsharing effectively requires careful planning and a robust framework. Let's break down its key components.
Defining Success: Metrics That Matter in AI
This is perhaps the most crucial step. Before any code is written or any data is analyzed, the client and consultant must unequivocally agree on what "success" looks like, and how it will be measured. For Gen AI projects, these metrics need to be specific, measurable, achievable, relevant, and time-bound (SMART).
For example, c2b recently created a 2 year program at a client, whereby if the Gen AI solution is an automated customer service chatbot, our outcomes were driven by:
Cost Savings: Reduction in human agent call volume by X%
Revenue Uplift: Increase in upsells/cross-sells through personalized recommendations by Y%
Efficiency: Average resolution time reduction by Z seconds
Customer Satisfaction: Improvement in CSAT scores by W points
Then, we measured Gen AI as a content generation tool, basing outcomes on:
Time Savings: Reduction in content creation cycle time by X hours per week
Engagement: Increase in user engagement (e.g., click-through rates, time on page) by Y%
Conversion: Improved conversion rates on marketing campaigns by Z%
The key is to select metrics that are directly attributable to the AI solution and are meaningful to the client's business objectives and are ACHIEVABLE.
The Baseline: Establishing a Starting Point
You can't measure a "gain" without knowing where you started. The baseline is the current state of affairs before the AI intervention. This must be meticulously documented and agreed upon. Using the customer service chatbot example, the baseline would be the current average call volume, resolution time, or CSAT scores. For content generation, it would be the current content creation cycle times or existing engagement rates. This baseline serves as the benchmark against which all future improvements (the "gain") will be measured. It removes ambiguity and provides a clear reference point.
The "Gain" Calculation: How We Measure Impact
Once the metrics and baseline are established, the next step is to define how the "gain" will be calculated. This is where the measurable impact of the AI solution truly comes into play. It often involves:
Quantitative Improvements: Comparing post-implementation metric values against the baseline. For instance, if call volume dropped from 10,000 to 7,000 calls per month, the gain is 3,000 calls saved. If each call costs $5, that's a $15,000 monthly gain.
Attribution Model: Ensuring that the observed gains are indeed due to the Gen AI solution and not other concurrent initiatives. This might involve A/B testing, control groups, or careful statistical analysis.
Timeframe: Defining over what period the gains will be measured (e.g., 6 months post-launch, 1 year).
The gain calculation should be transparent and auditable, agreed upon by both parties.
The "Share" Mechanism: Fair Distribution of Value
With the "gain" quantified, the final step is to determine how that gain will be shared. This involves negotiating a pre-defined percentage or formula. For example:
Fixed Percentage: The consultant receives X% of the calculated gain. (e.g., 10% of all cost savings achieved in the first year).
Tiered Percentage: The percentage increases as the gain increases, incentivizing even greater impact. (e.g., 5% for gains up to $1M, 10% for gains between $1M and $5M, 15% for gains above $5M).
Capped Share: A maximum amount the consultant can earn, even if gains exceed expectations significantly. This protects the client from excessively high payouts.
Floor/Guaranteed Minimum: A minimum fee for the consultant, even if the project underperforms slightly, to cover initial costs and basic efforts.
The share mechanism should strike a balance, offering significant upside for the consultant while ensuring the client retains the majority of the value generated.
Types of Differentiated Pricing Models for AI Consulting
While gainsharing is a specific form, it falls under a broader umbrella of differentiated pricing models, each with its own nuances and advantages.
1. Outcome-Based Pricing: The Direct Link to Results
This is a broad category where payment is directly tied to achieving specific, pre-defined outcomes. Unlike gainsharing, it doesn't always require a baseline calculation of improvement, but rather the successful attainment of a stated goal. For example, a Gen AI consultant might be paid X amount if their AI model achieves 90% accuracy in a specific task, or Y amount if the deployment results in Z% of tasks being automated. The focus is purely on the delivered result, with less emphasis on the journey or the degree of improvement from a baseline.
2. Performance-Based Pricing with Tiers: Scaling Rewards
This model builds on outcome-based pricing by introducing tiers of compensation based on the level of performance achieved. It's an excellent way to incentivize consultants to exceed minimum expectations. For instance, a Gen AI-powered fraud detection system might pay a base fee for achieving 85% detection accuracy, a higher tier payment for 90% accuracy, and a premium tier for 95% accuracy. This encourages continuous optimization and aims for higher performance benchmarks, directly linking consultant earnings to the quality and effectiveness of the AI solution.
3. Subscription-Based Models with Performance Overlays: Long-Term Partnerships
This model combines the predictability of a recurring revenue stream with the incentive of performance. Clients pay a regular subscription fee for ongoing AI services (e.g., model monitoring, maintenance, iterative improvements, access to Gen AI tools). On top of this, a performance overlay (which could be a gainsharing component or outcome-based bonus) is added, linked to specific achievements or ongoing improvements. This fosters long-term relationships, as the consultant is incentivized to ensure the AI solution continues to deliver value and evolve over time, making them a true strategic partner.
4. Hybrid Models: Blending Predictability and Performance
Often, the most effective pricing strategy is a hybrid approach. This might involve:
Retainer + Gainshare: A smaller, guaranteed monthly retainer to cover basic operational costs and initial development, combined with a significant gainsharing component for transformative results. This reduces risk for the consultant while still providing strong incentives for performance.
Fixed Fee + Outcome Bonus: A fixed fee for the initial build and deployment of an AI solution, followed by a bonus payment contingent on the solution achieving agreed-upon metrics post-launch. This balances the client's desire for budget predictability with the consultant's drive for impactful results.
Time & Materials (T&M) + Capped Gainshare: For exploratory or highly uncertain projects, T&M might be used for an initial discovery phase, transitioning to a gainsharing model once the scope and potential benefits become clearer. The gainshare might be capped to provide budget certainty for the client.
The key is to tailor the model to the specific project, its inherent risks, and the desired relationship dynamic.
Next time we will finish this series with a discussion of how we have successfully implemented gainshare pricing for our clients.
