What Is CU (Compute Units)? A Simple Guide to How Pimker’s CU Works

Question

Accepted Answer

Many people encounter CU (Compute Units) for the first time when reviewing Pimker plans and immediately ask: what exactly is CU? How does it relate to page count, execution frequency, and plan size? If you have the same question, that is completely normal. CU is not as familiar as traffic, article count, or keyword volume. But for AI optimization services, it is an important concept. Once you understand CU, it becomes much easier to see why different plans have different limits, why selecting priority pages matters, and how to choose the right plan for your business. This article explains CU in the simplest way possible. What Is CU? CU (Compute Units) can be understood as a usage unit for processing capacity inside the Pimker platform. In our FAQ, we describe it like this: CU ≈ AI processing workload per page × execution time However, for most users, the practical meaning is much simpler: No matter how large a page is or how much content it has, one execution = 1 CU. That means when you submit a page to the Power AI Engine for optimization, structuring, or submission once, the system deducts 1 CU. This makes things straightforward. You do not need to guess whether a page with more text or more images will cost more. Just remember this rule: 1 page × 1 execution = 1 CU That is the principle that matters most in real usage. Why Use a CU System? Some people assume CU reflects the exact technical resources used in the background. For example, a larger or more complex page should cost more CU. But in Pimker’s practical model, we do not recommend thinking of CU as a precise engineering metric. Instead, CU is better understood as an allocatable execution allowance. This gives users a major advantage: you can clearly plan which pages should be processed during each billing period instead of worrying about technical complexity. Most business owners, marketers, and website operators care more about questions like: How many pages can I process this period? Which pages should be prioritized? If I have limited CU, how should I allocate it effectively? So the value of CU is not in describing backend complexity. It helps you use limited resources on the pages that matter most. What Does “1 Page Executed Once = 1 CU” Mean? This directly affects your usage strategy. When we say one page executed once consumes 1 CU, it means: Homepage executed once = 1 CU FAQ page executed once = 1 CU Service page executed once = 1 CU A long page with many sections executed once = still 1 CU So the real factors that affect CU usage are not page length or complexity, but: How many pages you choose, and how many times you run them. That also means if your plan has limited CU, you should not send your entire website at once. You should choose strategically. Why Prioritize Important Pages If CU Is Limited? This is one of the most important concepts in CU planning. Because CU is limited, the best strategy is usually not to spread it evenly across every page. Instead, prioritize the pages with the highest business value, visibility potential, or conversion importance. For most business websites, top-priority pages often include: Homepage Core service pages Main product pages High-value FAQ pages Key conversion pages These pages matter not because they have the most words, but because they best represent your brand, products, services, and commercial positioning. So if your CU is limited, reserve it for these strategic pages first. In other words: Limited CU does not mean you cannot do AI optimization. It means page selection matters more. CU Is Best Thought of as “Slots” If CU still feels abstract, think of it as a number of execution slots. Each plan period gives you a certain number of slots. Your job is to decide which pages receive them. For example: 4 CU = 4 page executions 40 CU = 40 page executions Re-running the same page again consumes additional CU So CU is not just a technical number. It is a planning tool that helps you decide: Which pages are most worth processing this period? That shift in thinking is valuable because on most websites, only a portion of pages create the majority of business impact. Example: How to Understand the Starter Plan (40 CU) Let’s use a common example. Suppose you are using the Starter Plan, which includes 40 CU total. The plan is structured around one execution per week. The easiest way to understand it is: Starter Plan = 40 CU total. If executed weekly, you can select up to 10 pages at a time for Power AI Engine processing. Why 10 pages? Because the Starter Plan commonly runs over a 4-week service cycle: 10 CU available per week 1 page executed once = 1 CU Therefore, up to 10 pages per week Over 4 weeks = 40 CU total This makes the logic concrete. You do not need to think about model inference or backend computing time. You only need to know: If the plan runs weekly, Starter is ideal for focusing on 10 important pages consistently. Why Not Optimize as Many Pages as Possible? Many users assume that if they have many website pages, they should process all of them. But that is not always the most efficient strategy. Not every page has equal value. Some pages are secondary support pages, low-traffic pages, or pages with little commercial importance. If you spread limited CU too widely, you may dilute the optimization impact on the pages that matter most. Especially with smaller plans, it is often better to focus on the most strategic pages first. For example: Homepage = core brand signal Service pages = what you sell FAQ pages = how customers ask questions Contact or conversion pages = final action points Improving these pages first often creates more value than lightly touching many low-priority pages. Which Pages Should Be Prioritized First? If your CU is limited, start with these page types: 1. Homepage Usually the main entry point of your brand and a strong source for AI systems to understand your overall positioning. 2. Core Service Pages If you offer 2–5 major services, these pages are often more valuable than secondary content. 3. Main Product Pages For product-based businesses, prioritize the most representative and highest-converting product pages. 4. High-Value FAQ Pages FAQ pages are not just answers. They also help AI understand your brand, services, and use cases. 5. Key Conversion Pages Contact pages, booking pages, quote request pages, and buying-intent pages are also strong priorities. Choose a core group of strategic pages first. As your business grows, you can always upgrade later. When CU Is Limited, Strategy Matters More Than Volume Many business owners worry that smaller plans do not include enough CU. But for many companies starting AI optimization, the real issue is not insufficient CU. It is not identifying the right pages first. If your homepage, service pages, FAQ pages, and brand positioning pages are still unclear, even more CU may not create better results immediately. AI optimization is not only about quantity. It is also about priority. Ask yourself: Which pages best represent my brand? Which pages influence customer decisions most? Which pages deserve Power AI Engine attention first? Once those pages are selected correctly, even a smaller plan can create strong value. Upgrade Later as Your Needs Grow As your website expands, adds more service pages, more FAQs, or faster content updates, your original CU allowance may no longer be enough. That does not mean the smaller plan was bad. It means your business has grown. That is actually a positive sign. CU can also help indicate your business stage: If you only need a few core pages optimized, a smaller plan is suitable If you need broader content coverage, consider a larger plan If you want more pages continuously covered by Power AI Engine, higher CU will better match your needs CU Should Be Practical, Not Complicated The most useful way to understand CU at Pimker comes down to three ideas: 1. One page executed once = 1 CU Regardless of page size or content length. 2. If CU Is Limited, Prioritize High-Value Pages Use CU on your homepage, service pages, FAQ pages, and product pages first. 3. Example: Starter Plan = 40 CU With Weekly Runs = Up to 10 Pages at a Time A clear real-world benchmark for planning usage. So when you see CU again, do not overcomplicate it. Just think of it as: The AI execution allowance I can allocate to my most important pages this period. Once you understand that, CU is no longer just a term on a pricing page. It becomes a practical tool for smarter page selection and better resource allocation. If your CU is limited, that is okay. Choosing the right pages is usually more effective than trying to process too many pages at once. For most business websites, the pages most worth sending to Power AI Engine are not every page on the site—but the key pages that best represent your brand, services, and conversion value.

Learn AI