Last week’s post on RM stirred up some good response so I’m continuing with the topic (just in case I’m on some kind of roll here). 

I’ve noticed that there is no shortage of Reputation Management white papers available on the Web today.  Essentially, these “how to” white papers boil RM down to a three step process:

1.) Monitor
2.) Analyze
3.) Engage

In other words:
1.) Fly a small plane to Mars and collect the DNA samples of all Martians;
2.) Compare to human DNA, send results back to earth via satellite feed;
3.) Sit down with the Martians (after you’ve chased them around with needles trying to collect their DNA) and negotiate a WMD treaty.
(And please be back by Friday night; we have symphony tickets.)

What I’m trying to say here is that each of the *three* (ha!) steps is about as big as Texas and way more complex (I’m a Texan so I can say things like this and get away with it).  As always, the devil is in the details.

Let’s start with monitoring and analyzing.  You have, roughly, four options:

1.) Manual retrieval, sorting and analysis (or, Bandaids and Bubblegum)
This is where you assign a staff person (or nine) to search Technorati each day, sift through various feeds you’ve set up on multiple keywords, spend hours on forums and networks that are important to your brand, curse Andy Beal for putting together that nice list of about a bjillion tools you need to incorporate into your monitoring regimen, and then manually wade through the data and make judgments about what it all means.

Key Challenges:  If people are actually talking about you, (your brand, your products, your issues, etc.) the amount of data can become unwieldy PDQ.  If you have a PR crisis, you might end up drowning in data with no way to sort and analyze it quickly enough to respond adequately.  You will miss a lot.

2.) Semi-automated retrieval; manual sorting and analysis
A few tools out there will help you automate the retrieval process, i.e., collect data from monitored sources (Note: Be sure to ask what sources are monitored; I’ve seen “tools” that were no more than a lame attempt to recreate Technorati, with no monitoring of sources beyond the Blogosphere).  Automated retrieval can certainly help make the process more manageable, but at the end of the day, you still have to figure out a way to sort and analyze (positive vs. negative, new vs. existing, toss out the spam, identify influencers, etc.). 

Key Challenge: Not a lot of truly useful, “middle of the road” tools out there. You will still have to manually monitor some sources that the tool doesn’t pick up and you will still miss a lot.

3.) Automated retrieval, mostly automated sorting and semi-automated analysis
Companies like Nielsen Buzz Metrics and Cymfony offer the sticker shock solution, but if the volume of chatter is high and/or your needs are more complex, the price probably works out to be cheaper than the hours you would burn trying to manage it all manually (and these tools are certainly far more effective).  A couple of things to note/watch for if you go for door #3:  First, spam is a huge issue that no one has quite figured out.  This can impact your results dramatically.  Human beings are almost always required.  Second, even the cream of the crop solutions depend heavily on the brains of humans, so it is good to know who is looking at your data and making actionable recommendations (hopefully not an intern).

4.) Build your own: After all, even if you select an automated solution, there is always a fair amount of custom work that has to be done in order to set up proper monitoring.  Companies that build custom search tools are a good option (Vivisimo for example).  The World Bank recently released a beta version of a tool that is free to anyone who wants to participate in the open source Drupal community.

Some things to consider if you build your own: It takes a lot of computing power to spider a significant portion of the Web each day and bring it back for sorting (ask Google). Sorting that data by topic, positive/negative, etc. is a complex process that requires a significant amount of human brain power (ask me, I married a computational linguist).

The field will no doubt grow as agencies in the social media realm start to develop their own proprietary tools (us included) .  Of course, once you’ve got the monitoring and analysis part down, you have to do that “engagement” thing, which is another topic entirely (and many, many blog posts). 

(Note: We’ll be releasing our own white paper on Reputation Management at the London Search Engine Strategies conference in a few weeks.)
 

Filed under: Reputation Management on Wednesday, October 31st, 2007 by Nan Dawkins |
You can follow any responses to this entry through the RSS 2.0 You can leave a response, or trackback from your own site.