Topic: In the News! How much does a newspaper story affect hits?

[WeatherBeacon]

Two weeks ago I mentioned in another thread that a large regional newspaper ran a front page story about the Weather Underground that featured yours truly (http://www.wxforum.net/index.php?topic=1845.0). I mentioned that I would wait about two weeks to determine what impact the story had on the number of hits my site received. Well, it has been two weeks to the day, and it appears that the number of hits to my site has reached pseudo steady state, so here's my report for those who are interested:


We often wonder what impact a news story featuring our web sites might have on the number of hits our sites receive. Obviously there are many factors involved: factors that marketing experts and consultants spend a lot of time and money researching. I won't attempt to address those factors, but here's my story.

On Saturday, July 12 The Flint Journal ran a front page lead story about the Weather Underground that featured yours truly for local interest. Although the story's focus was the Weather Underground, the article gave me much more coverage than I anticipated. The story began on Page A1 in an 8-1/2 by 11-1/2 inch blue panel with lightning in the background. Inside it was a 7 by 5 inch photo of me at my station's sensor suite. The story ended on Page A2 inside a 7 by 7-1/2 inch panel. That’s a total of more than 1 square foot of prime journalism real estate on Pages A1 and A2! My web site's URL was listed twice: inside a 1.75 by 0.75 inch colored box on Page 1 and mentioned again on Page 2. The full story (with the same blue panels and photo) also appeared on two websites. Here's a link to one of them:

http://www.mlive.com/flintjournal/index.ssf/2008/07/kettering_mathematician_kevin.html

The full story ran again a week later in The Burton News newspaper, which is a section inserted into the Sunday Flint Journal specific to Burton subscribers. It included the same photo already mentioned, but also included a quote of mine in a large inset. The other difference was that this time the paragraphs were rearranged to highlight me (local interest story) before highlighting the WU. Then on Monday, The Burton News (the same section that appeared in Sunday’s Flint Journal) went out to subscribers of the Flint Journal in Burton. So the story appeared in full in three printed editions and also appeared in full on the newspaper's web site.

Needless to say, I never dreamed my site would ever get so much exposure!

A natural question is, "What is that paper's readership?" According to its own web site:

        The Flint Journal is circulated to homes in Genesee, Lapeer, Oakland, Livingston, Shiawassee,
        Saginaw and Tuscola counties. Daily circulation is 83,814 with daily readership of 174,620.
        Sunday circulation is 101,590 with readership of 227,679.
 
        Source:  Audit Bureau of Circulation: December 2004; Scarborough Research 2005

That's a lot of publicity for my site! So how did it affect the number of unique personal visits (PVs, not just page hits) to my web site?
 
In the month before the story appeared: 1) my PVs were down and I was averaging only about 10 unique PVs daily, and 2) on usaweatherfinder.com's hit counter, my rank fluctuated from around 102 to 106. After the appearance of the front page story, my rank rose to 28, and the number of unique PVs to my weather site went as follows:

  Day  Unique PVs   Conditions
------------------------------------------------------------------------
   0     199     front page A1 lead story appeared (Saturday)
   1      75     very nice, low humidity
   2      48     very nice, low humidity
   3      71     storms in area, WU down in evening
   4      91     hot, severe storms abound, intense lightning, heavy rain, small hail
   5      43     hot & humid, storms north and east but not here
   6      35     warm & humid but calm, no storms around
   7      27     normal temp & muggy, much rain, no storms (Saturday)
   8      27     normal temp & muggy, no storms (paper re-ran story in Sunday's edition-Burton section)
   9      28     normal temp & muggy, no storms (same section from Sunday delivered in Burton)
  10      32     normal temp & muggy, brief t-storm in early evening with light rain. Stronger storms further north.
  11      24     sunny, low humidity, beautiful day
  12      18     sunny, low humidity, beautiful day
  13      22     sunny, low humidity, beautiful day
  14      25     cloudy a.m., turned sunny after noon, low humidity, beautiful (Saturday)

There are factors that result in the number of unique PVs being undercounted. For example, as I understand it hits from my workplace, which comprises about 200 employees and a couple thousand students, count as only one unique personal visit per day even if many people visit my web site from different locations on campus. The same is probably true of many other offices and companies which connect to the Internet via one IP address.

So despite the newspaper’s readership of 200,000 give or take, despite the article’s being a lead story taking up more than half of page A1, and despite the article’s appearing in print three times and on the web, it seems that all that publicity has increased the number of unique personal visits to my site from roughly 10 to 25 daily. Nevertheless, my “15 minutes of fame” didn’t, in my opinion, result in a substantial (sustained) increase in visits to my site. You might say, “But the PVs more than doubled.” True, but relatively speaking that isn’t substantial in this case. For example, if PVs to my site had increased from, say 100 to 250 daily, that would be statistically significiant, but if PVs to my site had increased from 4 to 10 daily, that would not be statistically significiant. Likewise, seeing an increase from 10 to 25 PVs daily is not statistically significiant considering the amount of exposure my site received. So to those of you who regularly receive 50, 100, 200, 300,…, 600 or more unique personal visits daily: I tip my hat to you! How do you do it?

I think it was Chris (SLOweather) who has suggested that short but periodic citations in the news tend to produce more repeat visitors than one big story would. I suspect that’s probably true. Of course, some people probably just don’t care for my site. Perhaps people aren’t as interested in the weather as we enthusiasts think. Perhaps interest in the weather and in weather web sites depends more on demographics (economic status, educational background, upbringing, etc.) than we might think.

Nevertheless, I''m thrilled to have had the story! It has increased the number of PVs to my site about as much as I expected (but not as much as I had hoped). It will also give us something interesting to add to this year's Christmas letter.

Regards,

Kevin...

[SLOweather]

So to those of you who regularly receive 50, 100, 200, 300,…, 600 or more unique personal visits daily: I tip my hat to you! How do you do it?

Fine business on the analysis, Kevin!

As far as, "How do you do it?"... (These are general comments, not directed at anyone in particular's web site, and I have elucidated on most of them in the past, in various posts on various forums. And, they are my opinions only, of things I think have worked for me, and my observations of other people's weather sites.)

First of all, the most important thing is the your site makes you happy, otherwise, what's the point? So, if more hits makes you happy, read on. If, down below, I've thrown a rock at some feature of your site that you cherish, ignore me.

1) Have a compelling weather site to start with, complete with a descriptive localized domain name, not a free or ISP provided site. People here in San Luis Obispo can relate to "SLOweather.com". "Chrisesweather.com", not so much. Your URL's home page should open with current weather data, and/or the local forecast. A logo or banner is OK, but not a big one. No garish colors or graphics, no gratuitous animations, no spontaneous audio, no flashy Flash splash intro. Open on the weather where the user can see it. That's what they came there for. Look at your site through a first time visitor's eyes and see if you'd really return.

(The worst weather site I've ever seen was 15 screens deep on one page, had 2 overlapping spontaneous audio streams, no weather data at all on the top screen, multiple animations, large graphics, black background, repetitive information from multiple sources, and lots of non-weather content like gas prices and Amber Alerts. If you are putting up your own weather data, you don't need a Weather Underground sticker repeating it. And you probably need only one local forecast.)

2) The home page shouldn't be too long, 2 or 3 screens at most. And the top screen should entice people to scroll down to see the rest. One screen and a spiffy menu is probably better. People wanna know the current basic weather conditions and forecast. They don't care what kind of station you are using or anything static like that. Put the rest on other pages.

3) Use meta tags like keywords intelligently. Others will disagree with this, but my informal testing say it works. Put in your city name including common misspellings, ZIP codes, large local parks, colleges or universities, golf courses, weather, forecast, and anything special you provide (lightning, tides, etc). I get a lot of Cal Poly parents looking at the weather here where their son or daughter is going to school.

4) Promote your site as much as you are comfortable with. I've given out pens and left them behind in banks and stores. I advertised 3 months in a row in the high school newspaper last fall, with a school themed ad I created. I have a selection of Anole's weather graphics available and, a while back, I emailed a lot of local sites about them, offering them up for their sites and linking back to them, and offering to create custom ones. I probably have 10 other sites using them.

5) When a weather-related article or letter to the editor comes up in the paper, respond to it. Always put your website URL as a part of your sig, in plain text. Write your own letters to the editor. If you have a great story, send a teaser in. Master the news release for storm statistics, but make it timely.

6) Make sure that you have your contact info prominently on your website. I can't tell you how many sites I visit that don't have a way to contact the owner. If someone emails you, answer them courteously, and, if you have pens or other gee gaws, offer to mail them one.

7) Webcams and lightning detectors (if you can afford them) are like chum. But keep in mind that a webcam of something other than your driveway and the neighbor across the street is better. If you don't have access to a bitchin' view, then try a critter cam or a birdfeeder cam. Heck, I think people would watch an anthill cam.

8 ) Share links with other sites. This is why I started the Southwestern Weather Network a couple of years ago, and it seems to work.

9) Look for ways to add value or interest to your site. Add a weather blog, make sure the most recent posts appear on your home page, include current weather/nature photos, and ask for submissions from your users. Also, non-mainstream thinks like the Fire Danger and Watering Index scripts are examples of that.

10) (This one is a little new to me.) Add LOTS of weather content, even non-local. I've always made SLOweather a local weather site, reasoning that there are plenty of national sites from which to get other weather. But the current #1 site on USA Weather Finder ignores a lot of the suggestions I have above, and has a jillion pages. It obviously works for him.

11) Keep at it. I've been doing this since '98, and with the SLOweather domain since 2001.

12) Finally, get a decent hit counter to see how you are doing, perhaps one that uses cookies like Stat Counter. Kevin alluded to the problem of counting hits only by IP addresses, when you might be getting multiple different viewers from one IP address. USA Weather Finder tracks IPs, and my Stat Counter numbers are always larger. Stat Counter lets you create as many separate counters as you'd like, so I have different ones on most of my major pages. The free counters will be fine for most people.

One last thing... Bad weather makes good hits. Move to Tornado Alley.

[Anthony]

I to would like to improve my hits. Please take a look and evaluate my site. I'm open to suggestions. Remember I do not have the ability to run php or asp at this time. Even thou GD indicated I do. I assume it's the ad's that foul things up.

[jwyman]

Excellent writeup Slo! Thanks for the tips....

  

Jim

[SLOweather]

Excellent writeup Slo! Thanks for the tips....

  

Jim

You're welcome, Jim. I'm pleased that you find it useful. I'm by no means an expert on any of this, those are just my observations.

We might as well keep this going and anyone else can post here what they think works too.

Here's one I remembered just now.

13) On the one hand, your pages should be designed for your most common user resolution so they don't have to scroll horizontally. I can get that info from Stat Counter, and right now it's about even between 1028 and 1280. Remember to allow for vertical scroll bars etc. A 1024 resolution page should really be 945-955 pixels wide.

On the other hand, don't forget cell phone and PDA users. Make a simple (no AJAX/Javascript/Java) set of pages in that resolution, and link to that on your home page.

14) Got a GPS? Sponsor your own Geocache somewhere near you and label it with your site name, and/or make your Geocaching handle your site name.

[WeatherBeacon]

So to those of you who regularly receive 50, 100, 200, 300,…, 600 or more unique personal visits daily: I tip my hat to you! How do you do it?

Fine business on the analysis, Kevin!

Thanks, Chris!

As far as, "How do you do it?"... (These are general comments, not directed at anyone in particular's web site, and I have elucidated on most of them in the past, in various posts on various forums. And, they are my opinions only, of things I think have worked for me, and my observations of other people's weather sites.)

First of all, the most important thing is the your site makes you happy, otherwise, what's the point? So, if more hits makes you happy, read on. If, down below, I've thrown a rock at some feature of your site that you cherish, ignore me.

Please let `em fly!

.....

Great advice, Chris! Thanks for taking the time to write them out so thoroughly for us.

My web host provider has something called AwStats that we clients may use. When I click on it, a link that says, "Install AwStats on this domain" appears, so I have always been reluctant to install it. Is anyone familiar with AwStats? For all I know, I'm missing out on a great tool.

Thanks again, Chris!

Regards,

Kevin...

p.s.  A couple people emailed me asking how I got the paper to feature me. Reasonable question. The paper wanted to do a story about WU, so they contacted WU and asked permission to contact PWS owners in the area. WU won't give out a member's contact info without permission, so WU contacted me and asked if I would be interested in having the paper contact me. I said,..., well, you know the rest of the story.

[SLOweather]

I to would like to improve my hits. Please take a look and evaluate my site. I'm open to suggestions. Remember I do not have the ability to run php or asp at this time. Even thou GD indicated I do. I assume it's the ad's that foul things up.

Y'know, Anthony, I don't want to start being the arbiter of what's good and bad about any individual's site. I posted some general observations regarding sites I've seen over the years, and recommendations that seemed to have worked for me. If you think your site could be improved (I know mine is far from "perfect", whatever that is...), and want to use those remarks as a guideline, then you could perhaps print out that post, and use it like a check list as you review your site. Fix the things you want to, and don't mess with the rest. It's gotta make you happy, not me.

Chris

[saratogaWX]

Anthony,

Tom (Carterlake) posted more ideas along these lines on his site, and they're worth examining too

http://www.carterlake.org/buildtraffic.php

http://www.carterlake.org/weatherwebsitemistakes.php

I want to second Chris' recommentations, and also express his mandate to make your site so that you're happy with the expression of your weather data.  For two years, I had a one-page website (on Comcast) that showed the 5-minute summary of my weather conditions and a few thumbnails from Weather Underground.  It wasn't until I really wanted to dig into a redesign that I started adding stuff and quickly outgrew the capabilities of the free comcast website.  Then Tom (carterlake) and Larry (Anole) got me hooked on PHP and it's been a whirlwind since

The key thing is what do YOU want to see about the weather on your site?  What do you think visitors would like to see?
The blend of the two (and keeping it simple and maintainable by you) is what the focus should be .. it's a hobby afterall

Best regards,
Ken

[Anthony]

Chris & Ken,
Thanks for you input. Will probably just leave things as they are for now. Maybe someday I will be able to afford paid hosting and that will open up the ability for me to use php.

[jwyman]

Anthony,
  I was paying $9.95/ month for limited web space and NO php services. I ended up with ICDSoft and haven't looked back. Got a HUGE amount of diskspace, mail boxes, FTP users, huge bandwidth allowance, and the latest PHP and ASP support. And I am now only paying $6/month... And the 24/7 support has been fantastic...

Jim

[NiceBill]

Thanks fellers,

A great read, I am working on my web page and I have learned a lot.

Thanks again,

Bill.>>>>>>>>>>>>>>

[Anthony]

jwyman, I have considered using ICDSoft when I get the funds for paid hosting. But have also considered just sticking with GoDaddy for my hosting. Since that is where I purchased my domain.

[ironton]

My web host provider has something called AwStats that we clients may use. When I click on it, a link that says, "Install AwStats on this domain" appears, so I have always been reluctant to install it. Is anyone familiar with AwStats? For all I know, I'm missing out on a great tool.

OK, this is my first post here so hope I have the quote option done right....

Kevin,
I sent you an email from your web site link with some information on AwStats.  My host service uses this and i do find it useful.  I included a few screen shots so the email will be larger than normal.
-ed

[WeatherBeacon]

My web host provider has something called AwStats that we clients may use. When I click on it, a link that says, "Install AwStats on this domain" appears, so I have always been reluctant to install it. Is anyone familiar with AwStats? For all I know, I'm missing out on a great tool.

OK, this is my first post here so hope I have the quote option done right....

Kevin,
I sent you an email from your web site link with some information on AwStats.  My host service uses this and i do find it useful.  I included a few screen shots so the email will be larger than normal.
-ed

Thanks much for the info, Ed, and for the sample screen shots! It looks informative.

Also, welcome to a great forum! Enjoy!

Regards,

Kevin...

[WeatherBeacon]

Well, I generally do believe the mantra, "Bad weather equals many hits." (I believe Chris at SLOweather uses "Bad weather equals good hits.") Last evening the storms had little impact on bringing hits to my site, however.

Friday afternoon we were placed in a severe storm watch that stretched into the evening. From 6:00 to 10:30 P.M. we had training t-storms (none severe). The kicker is that we had cloud to ground lightning with a good stretch of steady torrential downpours. We received 1.54 inches of rain, most of which fell in a 2 to 2-1/2 hour period from 7:30-10:00. Nevertheless, after the storms began the number of unique PVs to my site rose from 21 to only 25, which is only 2 more than my daily average of 23 for the week.

The storms occurred in the evening after work and before people go to bed, so I would have expected the storms to bring more hits than that. Perhaps it's simply because it was a Friday evening and people were out.

Kevin...