How Often Should You Reprice? Choosing the Right Frequency

How often should a repricing system change prices – every five minutes, every hour, once per night? Few settings are underestimated as much as this one. Reprice too rarely and you lose visibility in price comparison engines and surrender the buy box. Reprice too often and you create system load, confuse customers and risk following every competitor markdown reflexively. Frequency is one of the central levers in dynamic pricing: it determines how fast your pricing strategy reaches the market – and how much control you keep.

What determines the right frequency

Four factors set the appropriate frequency, and they differ by assortment.

Competitor reaction time. The most important benchmark is how fast your relevant competitors move. If your three most important rivals change prices twice a week, hourly repricing is busywork for your system. If they react hourly, a nightly price run becomes a structural disadvantage, because your price is stale for most of the day.

Freshness of competitor data. Repricing frequency can never usefully exceed monitoring frequency. A system that recalculates hourly but receives fresh competitor prices once a day spends 23 hours recalculating on the same data. If you want to react faster, accelerate data collection first – in that order, not the other way round.

Margin and pricing headroom. Products with thin margins cannot absorb fast downward moves. The higher the frequency, the more important hard price floors and limits on step size become. A high-margin assortment can afford frequent small adjustments; a low-margin one needs rare, deliberate ones.

Technical and organisational limits. ERP systems, marketplace APIs and price comparison engines impose practical constraints: transfer intervals, API quotas, caching. A frequency your ERP cannot keep up with produces inconsistencies between shop, marketplace and comparison engine – exactly the outdated price displays it was meant to prevent.

Four frequency tiers compared

{% table %}

• Frequency
• Typical use
• Requirements
• Main risk

---

• Real-time (minutes)
• Marketplaces with buy box logic, electronics top sellers
• Streaming price data, high API quotas, tight guardrails
• Price spirals, high system and data costs

---

• Hourly
• Contested fast movers in your own shop and comparison engines
• Multiple daily price collections, automated rule sets
• Restless price image, customer irritation

---

• Daily (1–4 runs)
• Standard assortment of most online retailers
• Daily competitor data, maintained cost data
• Delayed reaction to sudden price drops

---

• Weekly
• Long tail, private labels, price-stable niches
• Basic market overview
• Gradual visibility loss when the market shifts {% /table %}

The tiers are not mutually exclusive. Mature repricing setups combine them: high frequency where the market is fast, low frequency where it would only create cost.

Differentiate by assortment segment

The most common mistake is a single frequency for everything. A proven split: the top 5–10 % of products with intense competition get the highest frequency, the broad middle runs daily, and the long tail – often half of all products – runs weekly. Promotional items need temporarily tighter cycles; price-fixed products and private labels without direct competition need no automated updates at all. Review the segmentation regularly: a product that climbs into the top sellers must also move up a frequency tier.

Case study: electronics retailer with 12,000 SKUs

A mid-sized electronics retailer initially repriced its entire assortment once per night. Its 800 top sellers were priced above the market for most of the day, and the buy box share on its main marketplace dropped to 41 %. The new setup followed the segmentation logic: hourly repricing for the 800 fast movers, four daily runs for the mid-range of 4,500 products, weekly runs for the long tail. After one quarter, buy box share rose to 68 % while the total number of price changes fell by roughly a third – and gross margin held steady, because price floors and maximum step sizes were introduced at the same time.

Common mistakes

• Choosing frequency by technical capability. Real-time is switched on because the system supports it, not because the market demands it.
• One frequency for the whole assortment. The long tail gets repriced as often as the top sellers, creating cost without benefit.
• Repricing faster than monitoring. Hourly runs on daily data simulate a freshness that does not exist.
• High frequency without guardrails. Every competitor markdown is followed within minutes – the spiral turns faster than a human can react.
• Set once, never reviewed. Competitive intensity shifts with seasons and market entrants; yesterday's frequency does not automatically fit tomorrow's market.

Summary

Repricing frequency is a market decision, not a system setting. Start from competitor reaction time, cap it by the freshness of your price monitoring, and secure it with price floors and step limits. Split your assortment into a few frequency tiers – fast for contested top sellers, daily for the middle, weekly for the long tail – and you achieve more market impact with fewer price changes than any single frequency could deliver.