Archive for December, 2007

On vacation: Japan

December 22nd, 2007 No comments

Arigato, you have reached Kelly’s inhousesem blog. I’m currently away from my desk in Japan until January 6th, but if you read some past posts below, and maybe waste some time reading celebrity gossip online, I’ll be back and posting before you know it.

Happy Holidays to All!

Categories: I Love the Interweb Tags:

Search and Acquisitions: What to look for in search when buying a company

December 21st, 2007 No comments

In the time I’ve been at Viator, we’ve bought one company, and we might buy more someday. Paid and organic search traffic are big revenue drivers for the company we acquired, as they are for Viator. When assessing the value of a company heavily dependent on search for revenue, some deep delving into the sustainability of their search programs as well as their additional growth potential is a must. Oddly, venture capital firms, and probably most companies, aren’t well suited to this task. I almost think a mini industry could develop; search engine marketing consultants who specialize in discovery for acquisitions. I’ve done search engine marketing discovery work for acquisitions a few times now, so here’s my advice on what to look for acquisition wise in search engine marketing programs.

  • Get access to every account, tool or reporting system in use. Download reports for the past year, by month. Download account listings. Thoroughly examine the settings for reporting and tools (cookies, attribution models).
    • No reporting is a bad sign. Not having an independent reporting source (ie, no Omniture type tool or relying solely on search engine provided reports) is not great, but its a step in the right direction. Ideally there would be a reporting system that provided unique transaction count attribution across marketing programs, based on last click or other sensible attribution rules. There would also be some form of organic search reporting.
    • Are the campaigns/adgroups, ads and keywords well organized? Do the URLs have tracking? How large are the accounts? Room for growth is a good thing. Having to redo everything because its a mess is not a good thing. Are the campaigns opted into content targeting? What are the general targeting and budget settings for the campaigns? Again, get a sense of growth potential – can you expand targeting or budgets to increase traffic?
    • This one is a little insidious, but run a change report in Google. Get a sense of how much active management is going on with the account. Also, have you run across any mistakes in the account?
    • Check for negative keywords. Every account should have some, if not, adding them will be an efficiency win.
  • Examine reports.
    • Are their inconsistencies with what’s been reported in other areas? Is the search marketing program growing over time and/or becoming more efficient? Look out for worrisome trends like increases in cost without increases in revenue. On the organic side, the trends are doubly important. Has position and traffic been sustainable over time and reliable or sporadic?
    • Is money being wasted on search? Look for really inefficient areas. Not tracking and optimizing content targeting separately is a common opportunity area where spend is inflated. Could more budget or distribution benefit search? Are the account bids regularly being optimized? Are there multiple ads on any adgroups or has ad testing been done in the past?
    • Check out all the public SEO info you can for at least Google and Yahoo – pages indexed, links, etc. Do they provide a Google sitemap or any other feeds?
    • Any accounts showing up in the reports that you haven’t been given access to or talked about? Its amazing what can be forgotten – oh right, we do have a paid inclusion account!
  • Talk to the people working on search and ask lots of questions. And this is key: take detailed notes. I’ve had to refer to notes about what people said or didn’t say during discovery, and you don’t want to rely on memory.
    • Just generally, do they know what they are talking about? Can they explain why the accounts are set up as they are and the trends in reports you have run? Have them walk through a typical day of work, are they doing what you’d expect?
    • Ask lots of questions about organic search strategies. Are they shady in their practices? Are the tactics sustainable? Have they ever been penalized by a search engine? If so why and did they learn from their mistake and resolve the issue? Be nice, but pry like crazy, the more information here the better in terms of coming to a valuation.
    • Where do they think the growth opportunities are?
    • This is just common sense – are the people or is the person working on search a good employee? Can you work with them? Are the forthcoming with information or do they seem to be hiding things? Vague answers are not a good sign, it often masks that they don’t really understand their programs. Also, the search marketer who pretends that everything is magical and way above your head is really, really obnoxious.

I prefer to get a brief overview of all the marketing programs and then dig into reports before having any detailed discussions. I have a lot more questions after checking everything out than I do before. The same kind of questions and discovery work also apply to email and affiliate programs, and even if search is a big component of a company’s marketing, you should ask about email and affiliates and dig in there as well, if they don’t exist or are small, that could be a big source of growth potential.

There’s probably more questions to ask depending on the marketing programs in use and the industry, but these are the common ones I run across. Do your homework!

Categories: Tip of My Hat Tags:

My latest functionality request – position by location

December 17th, 2007 No comments

I’d like to see the average position by location when I run a keywords report. That should be pretty easy to do.

Most of my campaigns target worldwide, or nearly so, for English language search queries. If I run a report to see how I am ranking, I get a worldwide rank for that keyword. But say I want to check up on how we are doing in the US or the UK, then I appear to be out of luck. I’ve come to realize that the overall rank is very deceptive. For example, say I have a keyword ranking 1.4, with great ROI, I am probably not going to bother increasing the bid since the position is nearly being maxed out, there’s probably no upside to gain. But that keyword in just the US could be ranking 5th, so there would be upside to raising the bid, I just can’t see it. Also, how efficient are we in the UK vs. US vs. Australia for paid search? Not sure. Is there a location we should just stop targeting for lack of sales? Not sure, though I bet if I dug around in Omniture I could figure this one out.

For an important keyword in an important market I use the ad diagnostic tool in AdWords to check on its rank. But that is hella not scalable. I could duplicate my campaigns and target them specifically to certain markets, but with twenty campaigns already, that is going to get unwieldy, as well as a maintenance hassle (Need to change a URL? Do it in six places now instead of one).

Ideally I’d like a filter so I can filter my report data by location, nothing fancy. According to my AdWords account manager, I am not the only person asking for something like that, so hope springs eternal.

Categories: Wag of My Finger Tags:

The Google Holiday Gift

December 11th, 2007 No comments

In one word: uninspired.

Who doesn’t already have some USB memory? Like maybe, the one Google sent me a year or two ago. And while the gift certificate is sweet, it was the present Yahoo! sent last year. Whoever is working on Google’s holiday gifts seems to have phoned it in. The card was pretty, but I am pretty sure the supposed to look like recycled cardboard packaging wasn’t post consumer.

Anyway all of this is just a reminder that ’tis better to give than to receive, so check out Viator’s Charity page. Yours truly picked the charities, with some input from the Viator staff.

Categories: Industry, Wag of My Finger Tags:

Matching to Less than You Think

December 5th, 2007 No comments

This is a lesson in why you should run Search Query Reports in Google. I noticed some jumps in spend and clicks for AdGroups for general destination terms, e.g., Rome Tours, which contains keywords like: rome tours, rome sightseeing, etc. I didn’t change any keywords, but everything is set to broad match, so maybe there’s some new queries that are matching. And thus, the running of the Search Query Report, solver of mysteries such as this.

I found that our keywords, in some cases, started matching to just the city name (e.g., rome). I think marketers mainly think of broad match as having queries match to more words than specified, not less. I don’t buy any city name keyword by itself. Maybe someday we’ll test these, but my gut says they are way too general to convert well, as well as fairly expensive and difficult to maintain a good quality score. Thankfully, this is easy to fix with some negative matching. I added “-[rome]” to our AdGroup, and now we don’t match to rome by itself, and rome tours, etc are still matching just fine. Hurrah for the negative exact match! Though why exactly Google thinks these more specific terms should match to a really general one is a whole other story.

Categories: Wag of My Finger Tags: