Archive for the 'Internet' Category

Amazon has just launched Amazon Widgets for their associates. Click here to see all of these widgets in action. Recently I wrote about BlogRush, which is coincidentally (or maybe not) a widget based system. Comparing the two companies, I feel Amazon’s array of widgets is really impressive.

Adding these widgets was really a breeze with the Ajax powered builder tools! So far I have only added a “product cloud” as you can see on the side bar.

Among all the widgets, “Context links” widget is most interesting as it will automatically identify and link contextually relevant phrases within your content to Amazon products.

A cool idea to monetize for bloggers and it is almost effortless to put a piece of javascript code near the end of my template just before </body> tag. The big winner will be Amazon as she taps into the masses of blogs out in the wild.

Will spend some time exploring the other widgets…

Related posts:

    TorrentFreak has an interesting article on how P2P is used as an marketing research tool. It is a good example of how to use technology to your advantage instead of fighting against it!

    Related posts:

      My earlier article on my web traffic experiment discussed about the usual approach of using social bookmarks such as Digg and StumbleUpon to get the initial wave of traffic.

      Today, I logged back into my Blogrush account just to poke around. It seems they met teething problems. The founder, John Resse wrote a “memo” to the members explaining several issues.

      I can understand the difficulty and challenges met by providing such a service. Apart from the network traffic, load balancers, app servers and such, I believe the most tedious work is to compute syndication credits for users and distributing the user’s link in an effective and efficient manner.

      Just to give an idea what kind numbers we are talking about… (The ratios are given by John)


      blogrushfigures.jpg

      The two tables above represent two simplistic scenarios. For the sake of simplicity, we assume each site has 100 impressions. For the first case, we imagine each user will only refer one other user. Think of it as a family tree, 1st generation would mean the user you referred directly, the 2nd generation would be your user’s referred user, so on and so forth.

      Of course, in real life, things may not be the case that one user only refer one other user. This particular case would bring the first guy (ie. you), 450 syndication credits. Erm… not really impressive considering the fact that there are 5 RSS entries in a widget, that would mean a 1/5 or 20% chance of being click if someone bothers to look at the widget.

      How about each user referring two other users? This would basically form what is known as a binary tree. The possible syndication credits now hits 27700 credits for this hypothetical example. Note that all parameters such as ratios, site impressions and generations are fixed, only number of referrals changed. That is the so call exponential effect!

      Again this is just a hypothetical example, in real world, the tree would be complicated thus making computation really interesting and challenging. Note that I am not here to discuss how effective this scheme is nor am I in the position to question the quality of these referral traffic.

      Each syndication credit merely allows your RSS entry to be shown in someone else’s blog widget. Some reader of another blog would still need to click, but at least it is unlike social bookmark sites where it is likely that users click together in a short span of time, bringing servers to a halt.

      If the reader were to search, one can find a myriad of comments that are both negative as well as positive. However, I would see it as less effort in the long round if one were to start some referrals now to enjoy the viral and exponential effect. Afterall, it is a game of numbers…

      For now, there are a couple of referral hits from the widgets but still low since I probably have no referrals. If you are a blogger and find my article useful, why not sign up with my referral link and give it a try?

      Related posts:

        I have been experimenting with ways to bring web traffic to my blog recently. Although computer engineering trained, I am no expert in traffic optimization but it does not take an expert to see that making money with blogs is basically a game of statistics, good contents and a myriad of other factors.

        I wanted to have fun with these experiments. The first experiment was with StumbleUpon (or SU). I came across SU from a blog community where people are into exchanging stumbles. Considering the facts that I do not write very often (being a weekend blogger), do not have much interesting contents and not a focused theme in my blog, I wondered if SU is useful to bring people to my site.

        plotstat.jpg

        Of course, those numbers are very pathetic in comparison with popular blogs but it was interesting to have these three spikes that occurred after I stumbled my own page. Those three entries were deliberately chosen, first is purely a single funny picture, second a record of my culinary experience (pictures and words) and finally an entry with most words.

        There is no rule in SU’s TOS that states that you can’t like you own blog entries (I may be wrong, please read their Terms of Services). It seems that stumbling own entry can drive the initial wave of visitors.

        Considering it is almost effortless to just click “I like it!” SU button on a page. It is really a cheap way to bring people to a page. In fact, the concurrent page view was too high for my server that had default settings (oops…) :P However, a general observation of SU visitors is that they come and they go, only a handful stay around to read more. After a wave of traffic, how then to measure retention?

        Before one can observe retention, it is a must to have some form of web analytics tool that monitors the site. Feeds are also another good way to monitor recurring readers. Feedburner can be considered one of the most popular tool for this purpose and best of all, the pro features are free!

        c.png

        Correlate the number of feeds subscribers from this feedburner stats graph with the earlier graph, we can see that the peaks matches around the same time.

        From the feed subscription stats, it is safe to conclude that the feeds subscription is quite dynamic, it seems subscribers constantly prune their feeds, removing stale feeds (like mine since I do not write daily). It makes sense, after all a RSS feed was meant to inform the readers of new entries, so if your site does not have frequent updates, why keep your feed? Simply put it, people like fresh new entries and news.

        As you can see so far, this is one of the strategies of getting web traffic. Obviously one must have reasonably attractive contents and a server that can support the incoming traffic. So it begins with ways to trigger the initial wave of visitors, after which, some of these visitors may become your feed subscribers.

        However, this approach of “initial wave then hope for subscription” is what I considered as an “uni-level” approach. What do I mean by that?


        unilevel1.jpg


        As you can see, between the tools and users, there is only one-level. Because of that single level, a writer has to constantly write, publish and publicized as part of the whole cycle. Of course, there are cases where the readers are bloggers too, and in some cases they hyperlink to your entry thus creating residual referrals.

        Whether or not that happens, really depends on your contents and readers. On the whole, this activity is a direct exchange of one’s time for readership, no constant writing and publicizing usually equals to no readers. There is little leverage that brings in traffic after the initial wave or effort.

        Having some background of Network Marketing as a developer who had previously done some projects to represent direct marketing networks and to compute their earnings, this trend of blog marketing is interesting.

        What many bloggers are doing, in hope of making $ from advertisements, are basically trying very hard to sustain or grow readership with this uni-level approach. There are other similar systems, like Pay-per-post, Hittail (it has residual properties if one writes according to suggestion) and alike but basically they are still doing the same thing, uni-level.

        Are there systems out there that are NOT uni-level? Will such systems help grow readership exponentially but not in big waves that will smash your server’s bandwidth to a grinding halt? Will such a system leverage on other blogger’s time?


        blogrush.jpg

        Today, I stumbled upon Blogrush. It supposedly is a system that syndicates RSS and has a multi-level effect. It is still in beta. I just signed up as part of my continuing experiment… Let’ see if it really brings in traffic.

        Related posts:

          Pro feature is free!!!
          Do subscribe http://feeds.feedburner.com/jymster ! :D

          Related posts: