GenAI products are far from their peak, most companies are still pricing ineffectively

S.T.G. — is a blog dedicated to the most important business thought leadership in Strategy, Technology and Growth.

TLDR:

1. LLM Pricing Shake-Ups: Small tweaks in pricing by big players like OpenAI can send shockwaves across GenAI products. Given OpenAI’s top-dog status, its pricing moves pack a major punch to the bottom line of GenAI companies.

2. Cost Over Cool: Many LLMs are good enough. Now, it’s all about getting the best bang for your buck. Watch as businesses hunt down the most wallet-friendly AI options to stay competitive.

3. Future of GenAI Pricing: Watch this space as GenAI evolves into savvier pricing models, switching between market-based and value-based strategies. High LLM API costs, though, mean freemium models are tough to profit from although data monetisation products might just be the freebie workaround.

Deeper dive:

At a high-level — there are a thee ways that products are priced (depending on the philosophy you take, market environment, and sales structure):

1. Cost-based-pricing — calculate the cost of production (how much you spend making the thing or service) and then add on a markup which is either a % increase of a flat rate increase.

2. Market-based-pricing — match your pricing similarly to other products on the market. Sometimes products are overpriced or underpriced for competitive advantage.

3. Value-based-pricing — you only make money when your customer is uses your product or makes a sale.

Typically, the way you charge customers is dependent on what is being sold. The interaction between what’s being sold and how pricing is determined is below.

There are a few key observations:

· Product-based goods are often cost-based interlaced with market-based e.g., construction which is the sum of raw materials, labour and additional costs depending on the thing being made and who it’s being constructed for

· Services-based good are almost always determined by the market — maybe because the output is intangible, and the cost is difficult to determine

· Data monetised goods e.g., social media, data platforms are often driven by how much value can be extracted

· There is a trajectory for data monetised goods to be sold on value to the customer or what the market values those types of interactions

· GenAI products are currently priced based on cost — the amount a typical customer may use the product and the cost of calling the GenAI API

So what are the insights?

1. GenAI products are massively dependent on OpenAI, Llama, Gemini pricing models. Even small changes in LLM provider can have huge implications for pricing of B2B and B2C GenAI products which will inevitably catch businesses out — especially while OpenAI seems to be head and shoulders above its competitors on many metrics and has somewhat of a monopoly

2. Cost reduction for LLM providers will soon become an important (if not, the most important) usage criteria for businesses. Many LLMs are already good enough to do many tasks that businesses need, now it will be about packaging the solution at an attractive price point

3. The GenAI market will change to a market-based or value-based pricing as the space matures. Other factors will determine if they take on a market-based pricing (like how Disney is priced to compete with Netflix) methodology or value-based (every time you succeed with using the product, the AI company will take a cut of your top-line)

4. It will be difficult to package GenAI products as a bolt-on (an additional feature you get with a pro version e.g., LinkedIn premium with GenAI insights for CV evaluation) because of the relatively high pricing involved with LLM API calls

5. For now, fremium GenAI products will be difficult to make profit from. It is likely that only data monetisation GenAI products will be able to become free for use because of LLM costs

Sources:

GenAI Needs Pricing Strategies to Match Its Potential (BCG): https://www.bcg.com/publications/2024/genai-needs-pricing-strategies-to-match-its-potential#

Pricing Generative AI Features for SaaS Products: Finding the Right Path (PSG Equity) https://www.psgequity.com/news/pricing-generative-ai-features-for-saas-products-finding-the-right-path

Three ways to get more value from your pricing process with generative AI (Simon Kucher) https://www.simon-kucher.com/en/insights/three-ways-get-more-value-your-pricing-process-generative-ai