Google Trends

Google Trends may be an interesting tool for investors.

  1. It allows investors to gather data on a company that’s fresher than the last quarter’s earnings release.  This can be helpful in turnaround situations such as Aeropostale (ARO) and Cafepress (PRSS).
  2. Having leading earnings indicators can be helpful for manufacturing companies (e.g. RGR, SWHC) where there is not much data on consumer demand due to fluctuating inventory at the retail and distributor level.
  3. In rare cases where fraud is suspected, Google Trends may provide some indications about actual revenues.

Here is an example of Google Trends in action:

google-trends-ARO

The lines on the chart are colour-coded to match particular years.  This makes it easy to compare search interest from year to year.

To the left of the chart, there is a bar chart for the average interest for a particular year.  Sometimes, this bar chart can be misleading if search interest follows seasonal fluctuations.  In this case, Aeropostale has higher search interest during the Christmas shopping season as well as back-to-school.

Some interesting Google Trends

ARO: It seems that the turnaround is not working.  For whatever reason, Aeropostale’s peers like ANF and AEO are also seeing their search interest and revenues decline.

PRSS: Cafepress is an interesting turnaround situation because the company’s founders have returned to fix the company.  The stock is arguably cheap because the company has sold off assets for cash.  However, the original core business seems like it may be faltering according to Google Trends.  I need to do more work on this company.

Some gun manufacturers (RGR, SWHC) look like they will see a weak Q1 2015 in terms of ultimate demand from the consumer.  VSTO‘s Savage Arms is bucking the trend.  Savage Arm’s recent resurgence may (or may not) be due to its $100 cash rebates.

Pandora (P) is interesting because Google Trends is indicating a rapid decline in interest while the competing Spotify service has seen dramatic interest growth.  Pandora’s SEC filings report high growth in revenues, active users, and listening hours.  There is a very big disconnect.  Because I have not done much work on Pandora, I do not know which set of numbers to believe.  I have no position in Pandora.

Nimble Storage (NMBL) has a small disconnect between its Google Trends data and SEC financials.  To be fair, its Google Trends data may be wildly inaccurate due to the small sample size.

Herbalife (HLF), Nu Skin (NUS), Coach (COH), and Lululemon (LULU) are some stocks in the headlines that have seen their search interest decline from last year.

Problems with Google Trends

  1. Search interest may not correlate with revenues.  For example, one-time events at Lumber Liquidators (a 60 Minutes segment) and Lululemon (see-through yoga pants) have caused traffic spikes.
  2. The trend may not be accurate, for the same reasons as #1.
  3. Useless China data because Google search is banned in China.
  4. If a company has multiple brands and product lines, you may need to aggregate data for searches from all of those brands and product lines.

Verifying revenue

I would absolutely love to figure out a method of easily verifying a company’s revenues where fraud is suspected.  Unfortunately, I don’t think that it can be easily done.  For example, the Crazy Eddie fraud took steps to avoid detection.  The company intentionally overpaid its taxes so that its numbers would not raise a red flag.  I believe that other frauds can overpay credit card processing fees, contingent rents, etc. etc.  They may also manufacture fake sales return data and other numbers reported in their 10-K that should track revenue.  To detect inflated revenues, having independent data is incredibly useful.  However, the problem with independent sources is that most of them are inaccurate.

US import/export data may be a good source of independent data.  However, I have not figured out a way to extract accurate data on the volume of goods imported by a particular company (without paying for a commercial service).

Google Trends can sometimes be inaccurate.  Other sources of web traffic data like Alexa, Compete.com, etc. can be even more inaccurate due to various problems with:

  • Methodology
  • Demographics (those who install the Alexa toolbar are a weird demographic of webmasters and people who install spyware)
  • Sample sizes (low # of Alexa toolbars installed).

So far, it seems that Google Trends is the least inaccurate source of independent data to verify a company’s revenues.

Other modes

Google Trends can also look at interest in particular countries.  For example, Michael Kors has growing interest worldwide.  However, its American search interest is roughly flat.

google-trends-KORS-geography

This might indicate that Michael Kors is close to saturating the American market (or may not be performing as well in the US).

(*I am long KORS.)

Appendix: a collection of some Google Trends results

Handbags

Guns

MLM

Luxury goods

Apparel

Post production software

Furniture retailing

  • Restoration Hardware (RH) plateaued in 2012/2013 and is in decline.  This is inconsistent with RH’s 20%+ revenue growth.  (*I am short.)
  • West Elm (owned by WSM) has seen its interest reverse and decline slightly.  This is surprising given West Elm’s rapid store count growth (19%) and comparable store sales (+18%).  West Elm grew revenues 26% while RH grew revenues 20.4% so one might expect both retail chains to have similar Google Trends interest.
  • Williams Sonoma (owned by WSM) shows rapid decline.  WSM has closed Williams Sonoma stores at a very slow rate (-2%).  Pottery Barn is also in decline.  It has been opening more stores at a very slow rate (2.6%).

Subprime retail

Retail

  • Tile Shop (TTS) is slowly growing.  (*I am not short for reasons explained in this post.)
  • Gamestop (GME) continues its slow decline.  (*No position.  I have been short GME in the past.)
  • Tractor Supply (TSCO) continues to become even more popular.  It looks like the company is continuing its momentum despite the change in CEO.
  • Ross Stores (ROST) is trending slightly higher than last year.
  • TJ Maxx (TJX) is gaining popularity faster than ROST.  Note that TJX has multiple brands.

Misc. technology

  • IBM is in a decline.
  • EMC has seen a very, very small decline.
  • NetApp (NTAP) is declining slowly.
  • Dot Hill (HILL) has seen its interest grow.  (*Unfortunately I am short and this may be a position I may close.)
  • Violin Memory (VMEM) peaked in 2013 and is now in decline.  (*I have shorted VMEM in the past.  No position.)
  • Nimble Storage (NMBL) has declined slightly year-over-year though it has grown over the past few years.  It may or may not be slowing down.  This seems to be inconsistent with Nimble’s astronomical revenue growth.  Small sample sizes may be causing inaccuracy in the Google Trends data for Nimble and Violin.

Malone complex – music

  • Sirius XM (SIRI) has slightly increasing interest though 2015 interest may be flat.  Sirius XM’s SEC filings report growing revenues.  (*I am long LMCA.)
  • Pandora (P) shows declining interest for the search phrase “Pandora radio”.  This is very weird because Pandora’s revenues, users, and hours listened are growing.
  • Spotify, a Pandora and Sirius XM competitor, shows rapid growth.
  • Live Nation (LYV) interest is in decline despite increasing revenue growth.  (I’d point out that fans may be more interested in particular artists’ concerts than Live Nation the company.)  Live Nation might not be doing a good job with its website and social initiatives.  Ticketmaster also has declining interest.

Malone complex – video content

Hot stocks

VMWare (VMW)

      has been declining since 2008/2009.  This is a head scratcher.  I would note that search interest in

virtualization

    peaked in 2009.

Consumer products

EDIT (6/10/2015):  One example of Google Trends tracking poorly with revenue is the online game World of Warcraft (publicly-traded ATVI owns a stake in the game).  Various websites such as this one summarize WoW subscriptionsHere is the Google Trends for WoW.

Advertisements

7 thoughts on “Google Trends

  1. Great post, I’ve been using Trends for a while.

    One comment though- remember the data is normalized in comparison with the total number of searches, so if total searches grew by 5% and a certain search term grew by the same rate, you will see a flat line.

  2. Re: Pandora. You need to remember that the vast majority of listening is driven through mobile (80%+), which is app based. If you want to use Pandora on your phone, you don’t google search “Pandora”, you use the app. Maybe the first time, you google for Pandora, to download the app, but from then on you use the app. This is why you need to be careful with any app based service – you actually need the number of new people downloading it to keep increasing vs overall Google Search volume (this is super high volume), which is hard to do as a business gets up the s-curve (but is still a ways from maturity). Also, you have the issue here with Pandora jewelry vs Pandora radio.

    • Thanks a lot for the comment!

      1- Pandora jewelry versus radio: I linked to the data for the search phrase “pandora radio”. That’s definitely something to pay attention to.
      2- The app explanation may explain the disparity between search interest and Pandora’s reported revenues. I didn’t think of that.

  3. Pingback: Egregious accounting fraud in retail | Glenn Chan's Random Notes on Investing

  4. The shift to mobile is changing search patterns. A site/brand can grow with search volume declining if people either a) go straight to the site without searching or b) have an app and don’t need to visit the site.

  5. Pingback: Ubiquiti Networks (UBNT) - Applying Apples Philosphies

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s