Because You're Worth the Rent

If you like living in an apartment and you don’t like paying money, this article is for you. 

We analyzed search trends to identify when the demand for apartments was the highest. We compared them with actual apartment prices to see what time of year is the best time to sign a lease. 

TLDR: demand doesn’t indicate anything by itself, sign a lease early in the year. 

We got 5 year keyword data from Google for the following keywords:

The only search term that has significant variation is ‘apartments for rent’. We’ll focus on that for now. 

So, let’s plot each of these over the months!

Clear as mud. 

Well, let’s not be too hard on ourselves -- we see some slight decline in demand towards the end of the year. This still isn’t enough for us - let’s dive deeper. Let’s get some real world data up here. 

Zillow has a public dataset that describes housing prices across apartments in many big and small cities across the United States. We took it a step further and separated cities into ‘urban’ and ‘suburban’ areas (labeled by our best friend, Chat GPT). Let’s see what we find:

WOW, HOUSE PRICES HAVE BEEN RISING!!

You didn’t need this article to figure that out. But what is more interesting is the different ‘waves’ in the curve -- they seem more pronounced in the urban trend versus suburban trend. 

Let’s do some more analysis, but we’ll make sure to account for rising prices over the years. We’ll focus on how expensive apartments are in each month when compared to other months in the year. 

And this is what we get. This is actually really interesting!

Urban apartment prices climb drastically until July or so, and then stagnate for the rest of the year. However, suburban apartment prices keep on climbing. 

For those who noticed: the y-axis are standardized values, and they don’t really mean anything in and of themselves. They are useful for ranking within a subset but not cross subsets- ie you can’t compare a standardized price of .5 for suburban apartments to a standardized price of .7 for urban apartments. These values are useful for comparing within a subset however, which is how we are using them. 

When combining our keyword analytics with real price data, we don’t actually find a correlation. Initially, I thought that this was driven more by a change in supply, as demand isn’t the only factor affecting prices. Upon research, however, I found that demand is typically lower during the winter months. Investopedia says “The lowest rental rates are usually found between October and April, particularly right after the December holiday season. Fewer people are interested in moving—the weather's bad, schools are in session, etc. So this is when you'll typically find the best rental bargains.” This directly contradicts our search data!

Why? Is Google wrong? Is it time to move to ChatGPT and sell all our alphabet stock?

In my opinion -- no. It’s just misleading. 

Let’s consider 3 scenarios, all college students:

Alice is graduating college next May, but she gets a job this November (yay!). She knows what city she’ll be in, so before she signs the offer, she’ll probably search ‘apartments for rent’ in whatever city she’s in to get a better idea of how much money she’ll need. 

Bob, in February, gets an internship starting in May (yay!). He started searching for ‘apartments for rent’ in February to get an idea of where he could live. 

Charlie is going through an exploratory phase. Two weeks before graduation (in May) (without a job offer), he decides to move to the Bay Area with this friend to start a startup (Good Luck, Charlie). He needs an apartment, so he looks up ‘apartments for rent’. 

All three of these people will sign a lease in May. However, when Google Trends registered their demand differs. This is the problem with Trends data -- people search for apartments wayyy before they sign a lease. Google Trends doesn’t pick up the nuance. 

So Google isn’t wrong (at least yet). It’s just misleading.

Also -- while writing this I had a question: what if ChatGPT did a similar thing with their queries as Google does with trends? Just some food for thought. 

Insights: