Alex Schultz: APIs

What’s going on with Linkedin?

According to all the data I look at regularly the second half of 2009 has really been a deviation from trend for them in terms of growth. They are doing a really great job. Does anyone have any idea why this is happening?

• email - the summary emails they are sending out now
• seo – the new business pages they launches doing great
• external – economy recovering a little, more jobs

I would love to get your thoughts. See the data below and links to the sources:

google trends data

quantcast data

compete data

focus = results

A year ago I was writing a blog post entitled “The Power of Focusing on Something” but in the end I couldn’t get it into great shape to post and so I chose to step back and leave it alone. Finally I am ready to post something a year on :)

Examples of losing focus for me

I have now prepared a few examples that make a lot of sense and I wanted to share. The great thing about having your own projects or your own business as a sole proprietor is you have no one to blame but yourself when something goes right or something goes wrong. My websites provide this for me. Below is a graph of traffic to my websites for the last decade (I left in the numbers since you can see them on quantcast here which measures things a bit higher than I do) in red and revenue per day in blue.

At every major life change in the above graph my website performance slipped esp. in terms of revenue (which is in my experience driven more by short term focus than traffic for advertising driven websites – sales, network choice and optimization etc…). Each of these changes was very necessary and good for me overall but the websites provide a good independent measure of how they distracted me from other things in my life. These changes took away focus and losing focus harmed performance.

Regaining Focus

I have settled into my job at Facebook now and so have some time to focus on my websites again. One thing I did recently was start to focus on videos of me making paper airplanes. These videos get a lot of positive feedback, are easier to follow than diagrams of me making paper airplanes and thanks to the new youtube revenue sharing also make me money. There is a graph below of the daily views of my videos on youtube.

You can see clearly what recent efforts I have made have done. Daily views on my videos have doubled and now something between 200 and 600 hours a day are spent by people viewing videos of me making paper airplanes in fact on average every visitor to my site views one paper airplane video.

My Conclusions

All of this spills into my real life. During those periods of negative website performance I also stopped going to the gym, stopped going out and really became pretty introverted.

So bottom line focus = results and lack of focus = fail. If you are going to make a major change as a business, as a team leader, as an individual, as a wife, husband, son, father, brother or whatever realize that it’s a trade off and make that trade off of focus explicit.

Is the upside of a new focus worth the downside of focusing less on something else?

dimensional reasoning, fermi problems

The above image has been burned into my brain for 7 years now. Not for the obvious reason, I am not horrified by the bomb or awed by the power or whatever (well maybe a little bit). This image and the story behind it taught me a great lesson about physics and business while I was studying for my undergrad at the Cavendish laboratory in Cambridge.

The lesson was that you can get to within an order of magnitude of calculating the power of a nuclear bomb using that picture and a bit of common sense. The full story can be seen here: http://www.tableausoftware.com/blog/visual-analysis-zeroth-kind-geoffrey-taylor-and-bomb and I highly recommend you read it. Spending a short amount of time and really thinking through a well defined problem should give you a good idea of the answer. If the order of magnitude of the result of a test isn’t interesting to you based on that reasoning, don’t do it. Move on to the next thing. Focus on the high impact. Your time on earth is short, don’t waste it doing trivial things.

The fermi problems are also an interesting space which I consider really similar to dimensional reasoning which is just using common sense to determine a value. My favorite is “how many blades of grass are there in a front yard”. The idea being you realize the yard is something like 10m by 10m and there are something like 20 blades of grass in a cm squared so you have 10^6 cm squared and hence 20million blades of grass in that lawn give or take (feel free to correct me if I got this one wrong). The idea is that it gives you answers that are correct to within an order of magnitude. You can read more about fermi problems at the links below.

http://infoproc.blogspot.com/2009/05/fermi-problems.html

http://radar.oreilly.com/2009/09/there-are-over-a-million-people-actively-using-facebook-right-now.html

http://en.wikipedia.org/wiki/Fermi_problem

I am a huge fan of doing this common sense checking before you do projects or to make sure that a number you are quoting is right. Really great reminder from o’reilly radar (an awesome blog) to think about this stuff.

big changes, big impact

Earlier this week I posted a blog post entitled “massive changes, minimum impact”. This blog post follows on from that. In making the navigation changes in that article which had essentially zero impact on the cocktail making site I also made changes to the advertising on the site. Moving the ads from the right to the left hand side of the page and moving them up 50-100 pixels.

Today I had my highest revenue day ever across all my websites and revenue has been climbing steadily since I made these changes despite user engagement staying flat and traffic volume growing.

Big impact is all about focusing on what you want. I have seen folks made huge impact in my internet marketing jobs over the last decade through minor creative changes in wording shifting text to be a call to action instead of a passive informational call out: in adwords, in onsite merchandising, in email and even in plain text links. Similarly I have personally made huge navigation changes (like the one I call out in the earlier article) with no major change to user metrics but massive change to revenue.

Plan your goal, make your change, optimize for your success metric and you’ll see massive success for minimum pain should you do it right.

Goodmail in my inbox – on unsolicited mail

I am confused. I thought goodmail was meant to fix spam or help it and yet here I have two goodmail certified mails from a company who I did'n’t subscribe to and have no idea who they are.

These are the first two goodmail certified mails I have ever seen.

massive changes, minimum impact

Last week I made a major change to my cocktail site. I moved the main navigation from the right to left and reduced the number of navigation links shown by half (to reduce duplication of top bar navigation with side bar navigation and give more space for ads). The before and after is below.

 

Before on the left and after on the right.

Below is a graph of page views per visitor before during and after the change from Google analytics. Can you spot when I implemented it? It’s the 3rd dot from the right. There was absolutely no change, I even studied the hour I pushed the change and nothing happened.

 

Sometimes things you think are important, aren’t. I stripped out half the navigation on every page on my site and completely changed the site and still people use the site the same way. I guess the advice is make bold changes, see what happens. Sometime’s it’s good, sometime’s it’s neutral, sometime’s it’s bad but the easiest way to find out is to try it.

My favorite dilbert campaign has something to say on this for internet companies and I reuse this all the time: http://dilbert.com/strips/comic/2008-03-16/

Quality work from traditional media in UK

Above shows high quality advertising and editorial work from Telegraph.

Below shows a story from the same day.

I 100% get the glass houses and stones argument but I find it really painful to see the (successful) digg spam attempts, headline grabbing, publish and be a weak imitation of a blog approach that newspapers seem to be taking in the UK. Then you couple with that absolutely caving on principle for the ads they are selling. On visiting the guardian earlier this year in the UK they seem to be the only paper who are not engaged in a race to the bottom in the UK newspaper industry. Perhaps that is why they are the biggest newspaper site in the UK.

using php to update status via the Twitter API - cocktail of the day on Twitter

You can now follow Alex’s Cocktail Making on Twitter @cocktaildaily so I’d love it if you followed me.

I have heard a lot about the twitter API lately being really really easy to use and so having hacked things together with the delicious, flickr, eBay, typepad and facebook apis in the past I thought it was time to give twitter a go. The whole development process start to finish took me 25minutes (started at 18:11 and finished at 18:36) from creating a new twitter account to having a daily updating script which pushed a random cocktail with over a certain score and a reasonable likelihood of not being rude to twitter. It took me 12minutes to write this blog post! Below are the resources that should make this easy for you to do and at the very bottom is the code I used (without my username and password)!

http://apiwiki.twitter.com/Things-Every-Developer-Should-Know (the twitter documentation, curl from command line at the bottom – it’s cool you can update twitter from command line)

http://morethanseven.net/2007/01/20/posting-to-twitter-using-php/ (this page showed me which curl-setopt I needed to send the message (it’s a post) and is now very slightly out of date but still should get you over the hump)

http://us2.php.net/manual/en/function.curl-setopt.php (the curl documentation)

For completeness here’s the code I used… you can easily modify it to do your own:

// Set username and password
$username = 'ENTERYOURUSERNAME';
$password = 'ENTERYOURPASSWORD';

// the message you want to send (I had a bunch of code here)

$message = “enter your message here”;

// The twitter API address
$url = 'http://twitter.com/statuses/update.json';

// Set up and execute the curl process
$curl_handle = curl_init();
curl_setopt($curl_handle, CURLOPT_URL, "$url");
curl_setopt($curl_handle, CURLOPT_CONNECTTIMEOUT, 2);
curl_setopt($curl_handle, CURLOPT_RETURNTRANSFER, 1);
curl_setopt($curl_handle, CURLOPT_POST, 1); // POST not GET
curl_setopt($curl_handle, CURLOPT_POSTFIELDS, "status=$message");
curl_setopt($curl_handle, CURLOPT_USERPWD, "$username:$password");
$buffer = curl_exec($curl_handle);
curl_close($curl_handle);

// check for success or failure
if (empty($buffer)) {echo 'there was a problem with the twitter updater';} else {};

Bloat vs Lite

A few colleagues here at Facebook put together a product called facebook lite. It launched last week to great acclaim (so well done to them) and really set me thinking about bloat on my sites (clearly pretty irrelevant sites to compare to Facebook but bear with me).

My paper airplanes site is where I have focused so far. All the site does is show you how to make 27 paper airplanes. There are video and diagram instructions and you can vote on something as your favorite plane. That’s it. Somehow I bloated this into 90 different php scripts to run the site ignoring include files and so on. No wonder I got hacked (it happened again in February and I failed to notice) and always felt sick when it came to planning to update the site.

I dug in a bit deeper since a lot of the components had a tonne of emotional value to me as every single bit of the site was built by me from scratch and really it was this site and each of those features that helped me get into being an internet marketer.

Videos

I specifically created a section to make my users view more pages using video: for each paper airplane on the site I created a second page that contained an embedded video from YouTube rather than just including the video at the top of each paper airplane. I have reversed this decision now making it better for the users even if less page views for me. At this stage with all this burying still 3000 people a day are viewing these videos, they are better than the diagrams. This is my most compelling feature and is buried.

Unusual/Fast/Alex’s Favorites/Gliders

Together these made up <10% of my page views and yet were linked from the main pages. They initiated <5% of my visits and when I dug in it was just because from time to time they got ranked for the word “paper airplanes” instead of planes.php as the second, indented result from my site. With all the link love and real estate they get on my site if they can’t do better they are gone.

search

On every page on my site I link to a search product from atomz. wtf?! Initially I thought I had better install Google co-op then I thought I’d look at the data first. In the last year this page was viewed a total of 1000 times. What a waste of valuable real estate. On cocktailmaking.co.uk everyone needs to search and with a similar link 30-40% of my visitors use the feature. On paper airplanes with a total of 27 pages browse is easier and when presented with the option, no one uses it. Gone.

index.php

This had 20 links to different sections to browse paper airplanes as navigation, most of these sections are now gone. Now it has one large link to browse more planes and an embedded video to start viewing the planes on the homepage itself without going any deeper. I tried replacing it with the planes index itself and decided I didn’t like how it looked. I will ab test this in the coming weeks and share the results here.

paperairplanes.co.uk does one thing and does it well: show you how to make a paper airplane. It doesn’t take 90 php scripts to do that. It takes one page for each paper airplane and a couple of navigation pages. That is it.