Welcome to the Q&A Forum

OK cool that's reassuring, thanks!

The title tag for this page is the following:

"Best Mothers Day Gifts, Unique Mothers Day Gift Ideas | UncommonGoods"

However, when you search Mothers Day Gifts UncommonGoods the title tags show up differently. What's up with this? Can I fix it? Is it a problem?

Would you rather have someone +1 your homepage or add you to their Circles? Which one will help you out more in the SERPs ?

Thanks for the feedback. You guys are right about Youmoz. I just submitted it now after adding some screenshots, more thorough directions, etc...keep an eye out for it.

-Zack

A few weeks back, I went to a Distilled meetup here in NYC.  SEER Interactive's Mark Lavoritano did some cool slides on the seasonality of keywords. Basically, his presentation made the point that you should not only think about which keywords you want to rank for but also WHEN they are most valuable.

This made me think...we have a lot of moving parts to our marketing efforts. Emails with interchangeable modules, a homepage with interchangeable links, and other dynamic elements for which we have to decide what themes we want to market for the week. Babies or bikes? Kitchen Gadgets or Wine Glasses?

Google Insights for Search  is a great tool which allows you to look at keyword traffic year over year. However, for many of the keywords (like the ones mentioned above), on a multi-year timeframe, it can be tough to sift out the specific weeks in which traffic repeatedly peaks year after year. What I really wanted to see was the last 5 years laid on top of each other to find the common peaks. Even better, if I could map 5 years of keyword data to a single row in a spreadsheet and then use conditional formatting to create a colorscale, I could create a sweet forecasting calendar with several keywords and use this to choose the best timing for various marketing campaigns. Here's a link to a screenshot of the calendar I created: forecasting calendar

I could have done this in excel, but I've been wanting to try out Python for a while now and decided this was a great time to do it. After some reasearch, I figured out how to import a csv into python and the rest was done with for loops and lists, which is fairly basic python. I've pasted my code below. In a nutshell, the program runs through all 5 years of traffic data and increments a count in a list whenever it sees a peak (according to a threshold called "peakInterestValue" that you set in the code). The output is a list of 52 numbers [0-5] (representing 52 weeks over 5 years). If the value is a 5, it means that all 5 years showed a peak in traffic at that week. If it's a 4, then 4 (out of 5) years showed a peak that week, etc...you can then copy/paste this to a row in an excel sheet with all your keywords, apply a color scale w/ conditional formatting, and  boom! you've got a forecasting calendar.

This code works on the exact file that Google insights exports so you don't need to format it at all. It's ready to rock.

If you want to see the code formatted and cleaned up, check it out here. If you want to see a hot mess, I've also pasted the code below. You can drop it right into a .py file and run it off a cmd prompt but you'll need to install python first:

http://www.python.org has installation info and great tutorials as well.

Enjoy!

import csv """reads a file from google insights""" """Open the last 5 years of data from Google Insights""" anniversarygiftFile2007 = csv.reader(open("anniversarygift2007.csv","r"))anniversarygiftFile2008 = csv.reader(open("anniversarygift2008.csv","r"))anniversarygiftFile2009 = csv.reader(open("anniversarygift2009.csv","r"))anniversarygiftFile2010 = csv.reader(open("anniversarygift2010.csv","r"))anniversarygiftFile2011 = csv.reader(open("anniversarygift2011.csv","r"))

"""Combines the data into a list"""

anniversarygiftFile = [anniversarygiftFile2007,anniversarygiftFile2008,anniversarygiftFile2009, anniversarygiftFile2010,anniversarygiftFile2011] """counters"""i=0j=0 """flags used to initialize lists"""

definedFlag=0

definedFlag2=0 for i in range(0,5):

j=0

for row in anniversarygiftFile[i]:         if j<=4:

"""skips the first 5 rows"""

elif j==5:

"""initialized the list on the first week of data"""

anniversarygift=[row[1]]

definedFlag = 1

if(i==4):

peakInterestWeeks=[0]

else:

"""appends the list with each row"""

anniversarygift.append(row[1])

if(i==4):

peakInterestWeeks.append(0)

if len(anniversarygift)>=52:

print("i = ",i)

if (i==0):

if(definedFlag==1):

anniversarygiftArray = [anniversarygift]

definedFlag2 = 1

elif (definedFlag == 1):

if(definedFlag2 == 1):

anniversarygiftArray.append(anniversarygift)

break

j=j+1       i=i+1

""" Now all of the data is in python lists""" i=0

j=0

""" Lower peakInterestValue to lower the traffic threshold and discover more peaks """

peakInterestValue=90

""" This is a variable to help you tweak peakInterestValue"""

peakInterestCnt = 0

for i in range(0,5):

print("i =",i)

for j in range (0,51):

if int(anniversarygiftArray[i][j])>peakInterestValue:

"""If keyword interest peaks, peakInterestWeeks[] is incremented"""

peakInterestWeeks[j]=+=1

peakInterestCnt +=1

print("Peak interest",peakInterestWeeks)print("Peak Interest Count =",peakInterestCnt)

"""peakInterestWeeks[] is printed out to a row in an excel file"""

c = csv.writer(open("anniversarygift.csv", "w"))

c.writerow(peakInterestWeeks)

Ok thank you. I will email directly.

Hey Carin,

I just wanted to follow up on this...I'm still seeing these spammy binary files show up as links. Unfortunately it makes OSE quite useless for me in regards to exploring our own backlinks.

What is the status of this problem? Has there been any headway ? Why does our site have problems but most others don't?

Thanks!

-Zack

Wow. i had not seen that yet. Def going to read that

This is awesome feedback and IMO tough love is always best..thanks for taking the time to give me honest critique.

In response to your questions..

1. What part of this article is "linkbait"?

We hoped that identifying angry cities would be controversial enough to be "linkbait"

2. Who is your target audience?

We have no target audience. Great point.

3. How did u come up with this topic? The foursquare rudest cities post was the inpiration...how do you think ours compares?

http://engineering.foursquare.com/2011/02/28/how-we-found-the-rudest-cities-in-the-world-analytics-foursquare/

4. Yes I see what you mean. Im going to discuss with our group..

we just relaunched our blog with a new design so we still need to add sharing buttons.. u are absolutely right on this.

Your comment on topics is very interesting...do you have any resources you could suggest for topic popularity research ?

Thanks again for this extensive feedback, much appreciated by all of us!

This is good, I think I see what you're saying...so a few things here to clarify:

Regarding "chest thumping" and "marketing"....do you think we need to start off with a more neutral point of view? Less biased?

Regarding the BS...it was included as a joke meant to soften the tone and show that we don't take this data to seriously since F-Bombs don't necessarily equal angry people...Do you think that it would be more effective it was written in a more objective, serious tone ? My thinking was that the data is not compelling enough to be linkbait on it's own and i thought we needed humor to make it worthwhile. Am I wrong? Do you think that the data is interesting enough on its own without all the BS jokes and random city facts ?

I work at Uncommongoods.com. I ran a big linkbait brainstorming session a few months back. Several ideas came out of the meeting and we just released our first piece of "linkbait" that was conceptualized during the brainstorming session. I think it's safe to say this first piece of linkbait is a big fat failure as it has only received 2 likes and no comments in the past few days. I want to know what I could have done to make this post more compelling so our next post can be better.

http://blog.uncommongoods.com/2011/kindest-vs-angriest-cities/

Any thoughts ?

One of our big natural search competitors for gift keywords is Gifts.com. We are competing for many keywords like "teen gifts",  "gifts for him", "gifts for her". For many of these, the Google SERP has multiple Gifts.com pages on the first page.

I have never seen more than one of our pages (uncommongoods.com) on  a SERP page. Any clue how/why Gifts.com has multiple pages in search results ?

Thanks!

I recently submitted an XML sitemap on Webmaster tools:

http://www.uncommongoods.com/sitemap.xml

Once Webmaster tools downloads it, how long do you typically have to wait until the pages index ?

I’m starting to notice some interesting behavior in Google rankings. It seems like most days, rankings move around but movements are relatively small. Then, maybe once or twice a month, there are days (like today) where everything seems to shift around in more dramatic fashion. I don’t know if it’s algorithm updates, or maybe they do light crawls every day and heavy crawls a few times a month..?? anybody have any idea what is behind these big waves of ranking changes?

Hey Carin-

Thank you so much for this in-depth response. Glad to hear that you guys are aware of it and trying to sort it out.  Very interesting info...I'd never hear of "binary" links before but I hope you guys can figure out how to handle these. Seems like a tough task to tackle, just by looking at my CSV it looks like these come in several different forms and they could be hard to identify..I have a few questions:

1. Is there by chance a URL you could give me that points to the old OSE ?

2. How often does OSE crawl? Is it a constant process or are there scheduled crawls?

Thanks!!

-Zack

OK cool good info, hope they fix it soon!! Any good ideas on how you can filter this crap[ out ?

I am starting to miss the old OSE. I've found that for a lot of the pages on our site, the new OSE is showing WAY more links and most of them are garbage nonsense links from China, Russia, and the rest of the internet Wild West.

For instance, in the old OSE, this page used to show 9 linking domains:

http://www.uncommongoods.com/gifts/by-recipient/gifts-for-him

It now shows 454 links. Some of the new links (about 5 of them) are legitimate. The other 400+ are garbage. Some are porn sites, most of them don't even open a web page, they just initiate some shady download. I've seen this for other sites as well (like Urban Outfitters) This is making it much harder for me to do backlink analysis on bc I have no clue how many "Normal" links they have. Is anyone else having this problem ? Any way to filter all this crap out ? See attached screenshot of the list of links I'm getting from OSE.

NHXnn

Yeah unfortunately they aren't ready to manage an additional DB

No that's definitely helpful thank you! We'll make sure to not use them

I'm trying to convince my tech dept to take our blog off of a subdomain and move it to a folder on our root domain. Understandably, they don't want to build it from scratch, so I need some good CMS tools to build the blog.

Our first choice is Wordpress but it isn't compatible with Oracle. Does anybody know of any good solutions that are Oracle-friendly? Our platform is PHP..

Thanks!!