I recently posted a study on lockboxes in the Charlottesville area on our Nest Report site. We keep that information up to date all the time in house. But we don’t release it very often in the public realm. The reason for that is that sometimes information exists that is hard to interpret, and which can include inaccuracies. This is one of those areas. We like our lockbox data, but I want to explain the limitations and potential errors.
First and foremost, these boxes are great, and I can’t imagine my job without them. In terms of functionality, they rank super high in the customer satisfaction ratings. If you ask agents if they like them, they probably don’t even think much about them, because they just plain work… almost all the time.
A primer on how these boxes work. There are two parts to the lockbox, There is the lockbox itself which stores the actual key for the house and stays attached to a doorknob or rail. The other half is the actual key that agents carry with them. When an agent arrives at a listing, a passcode is entered into the key and it communicates via IR technology to the lockbox which subsequently opens. The key then stores the lockbox serial number in its memory along with a time stamp. Each evening, the key must communicate via either computer or phone line with GE who takes that serial number and time stamp and alerts the listing agent of who showed which house and when.
Then, Nest collects that data from GE and puts together the trendline info you saw in the article. But there are some real limitations to the information.
Agent cooperation: The key that is used to open the lockbox is sometimes a smart phone, and sometimes a little black box like the one pictured here. When it is a phone, no problem. The phone automatically updates pretty much instantly that the house is being shown. But… if the little black box is being used, the agent needs to take some action for the upload to occur. Let’s say an agent shows a listing, but doesn’t go out for another week or so to show more clients. That agent doesn’t NEED to update their key, and so the info takes as much as a few weeks to update. The listing agent finds out 10 days later that their home was shown. And our data gets updated 2 weeks late.
Workaround: We looked every week for several months at the weekly data. We compared Week X data at the end of the period, a week later, two weeks later and so on out about 2 months or so. What we found is that at the end of the week, as few as 60% of showings are accounted for. BUT, within 2 weeks, roughly 95% of showings are in the system. So, we just ignore the most recent two weeks. We start paying attention about 14 days out. As more and more agents are going smart phone, the time delay is going down.
Agents opening boxes, but not showing property: You may not think this is a problem, but it is. Remember, we can’t see what boxes are being opened, or by whom, or why. So, an agent who is listing a new property may very well be in his office and open a lockbox to check that there is no key inside. He then may go to the house and open it again to insert a key. He may then go back with a photographer, or a contractor, or a floor plan artist. The house still isn’t listed, and a lockbox associated with it has been opened 5 times.
Now, think about the other end. When a house sells, typically a buyer looks at a house multiple times, then puts in an offer. After under contract, the buyer goes back for at least one inspection. Plus a contractor to price repairs. Plus the termite inspector. Plus the listing agent getting the key out. Again, this is 4 openings of a lockbox that do not count as a showing. All AFTER the contract is written.
How about the listing agent. On a vacant property, a listing agent might check the property as often as once every other week. Again, openings that don’t count as showings.
So, what is the trendline good for? Lots. All these openings that don’t count… They provide us with a baseline. Even if no one shows a single house in all of Charlottesville, there are still going to be some openings. You just get used to this. So what we look at is not the actual number of openings, but the change in that number from period to period. We are looking to see if showings are up or down from last week to this week and from this time last year to this week this year.
But, because there is a baseline, and we can’t know EXACTLY what that number is, it means that a 5% shift is really MORE than a 5% shift. Up or Down.
Take the change from the weeks of Sept 1st to the week of Sept 8th. The lockboxes showed a decrease from 1390 to 1299, a decrease of 6.5%. BUT, if the number of those openings that wasn’t really a showing was say 400 in each period, then the real shift is from 990 to 899. And that difference is 9.2%. However, it is easier and more consistent for us to just use the actual numbers than for us to try and determine what the “true showing value is” plus… we’d never get it right.