Scrape vs Actual Rate Differences

Why Scraped Rates Might Not Match Actual Rent Off Airbnb, Vrbo and OTAs? 

When comparing scraped rates against actual rents on reservations, there is typically a small median error that many people look at in first glance with a "this scraper must be broken". It is important to understand the nuances between scraped data and actual reservations to understand this better. The best scrapers in the world have lower median errors but are never 100% accurate. There are a few great reasons for this: 

Fees and Channel Mark Ups & API Differences

One reason why scraped rates might not match actual rent is because of fees and channel mark ups. Different OTAs charge different commissions on bookings, and hosts typically have their own fees on top off. Managers will often roll channel mark up fees into rent to help cover the costs of the channels, a strategy called net rates. This may cause the scraped rate to be higher than the actual rent when these fees are unbundled. 

Also along the fee bundling lines as mark ups are the other fees managers charge. Depending on the Property Management Software (PMS) and the API of the OTA certain fees may not be lump sum and dumped into the cleaning fee bucket. Some fees may be daily or percentage based in which case they may be dumped into the rent bucket, depending on channel, and backed out later. This will vary by manager, but ultimately there is a upper bound to these fees as guests and owners will start to revolt. These fees are later unbundled, similarly to the channel mark up costs. 

Not all managers participate in mark up fees and less managers will be bundling fees in this manner. The bundling and mark up require a fee and channel strategy as well as the technical understanding to auto pull these out as they come back into the PMS. Because of this when evaluating in aggregate the impacts are relatively minor. Working with a reputable and knowledgable scraping coming can help you account for and understand these differences. 

If you are building a pricing algorithm, the most important feature is modeling the relative shape of the curve, picking an elevation and then adjusting based on demand signals. So a good pricing algorithm will be able to deliver superior revenue performance with no measurable impact. 

Lags in Scraping 

Another reason why scraped rates might not match actual rent exactly could be due to lags in scraping. When you scrape rates, you are essentially getting a snapshot of the current rates. However, these rates can change rapidly, especially during booking extremes (low or high) or when a manager is aiming to drive more last minute demand. There may also be managers without a strong channel router that may not update all channels as quickly as needed to minimize this impact. 

Selecting a good scraping partner who scrapes listings frequently will help to mitigate this impact. At Hungry Robots our service hits listings anytime there is a change and logs to help mitigate these impacts. You can't have a one size fits all rule to scraping frequency due to the individual revenue management strategies or levels of expertise out there. 

Quoting Differences

The individual quote may differ from the average price. What we mean by that is a manager can have a different price based on the length of the stay which may trigger additional discounts or minimum pricing requirements. If the reservation is over 7 nights or hits the managers triggered discounts the reservation rent will differ from price. Managers can even have the option of itinerary based price, which means a different rate for every permutation of check in and check out dates. A manager may chose to do this if they are trying to sell harder to sell dates or if the reservation will create gap nights or a back to back clean. 

Quoting differences can be difficult to mitigate but often are within a median error that will be tolerable for most analysis. It is also important to note the number of managers with the sophistication for itinerary based pricing is limited and on aggregate the length of stay discounts will be within a tolerable range. 

If however you want to mitigate this impact and have much more accurate scrape rates, it is recommended you work with a scraping provider who can offer quote based scraping to deliver not just one price per date but the full array of options from that date into the future. Hungry Robots has quote price delivery if you need this level of accuracy. Reach out and schedule time to discuss! 

Other: The most annoying category! 

Things happen! Especially in short term rentals which have to deal with level of amenities, near by construction, home cleanlieness, differing expectations, natural disasters and a million other things. Because of this endless list of factors and a screaming guest on the phone, customer service agents may have limited tools to react to a situation. Guest service agents may change rent to account for refunds, credits, technological issues, quote mismatches and more. 

Luckily these issues are the exception and not the rule. On aggregate these are marginal and single or fractional percentage points. They are never zero so you should keep this in mind. 

Conclusion

Due to all of these factors it is important to work with a expert data scraping service with STR direct experience so they can mitigate the factors that can be controlled for and help you understand the ones that cant. 

Looking to learn more? Schedule time with Hungry Robots so we can help with your data needs and how to best approach your use case. With our extensive STR experience building data products and STR companies, we can help you with your competitive edge.