Sunday, September 30, 2007

SearchQuant Top 5 All-Time Posts

In the interests of saving you all the time of going into the SearchQuant archives for The Best of SearchQuant, I've created a list of my top 5 posts of all time (as measured by views):

#1 - Expanded Broad Match: Taming the Beast
#2 - A Brief History of Bid Management (YouTube video)
#3 - F#@k Click Fraud, It's Distribution Fraud That Matters
#4 - Does SEM Cannibalize SEO? Not Likely In The Real World
#5 - Get Off My Portfolio Algorithms

*NOTE: this post is serving a dual purpose, dual meaning I'm running a Facebook Flyer CPC-based ad on the Facebook networks for Yahoo and Google employees, basically to see what kind of CTR I can pull within Facebook's CPC-based ad format.

Friday, September 28, 2007

Google Radio Ads- poor results so far

I've been trying to follow advertisers' use of Google Radio Ads (the result of their DMarc acquisition), and haven't yet heard of anyone having other than poor to ambiguous results. There are two good, thorough reviews on DigitalPoint Forums from people who've run Google Radio Ad campaigns in the last months. Exceprts:

Goramba says:
"The first test was done on a budget of about $1,000 for 30 days. I selected 4 major cities, requested it NOT play after 12am, but had no other restrictions. The effect: WORTHLESS! After playing for about a week Google claimed 293 plays to 201,000 listeners. Watching my Analytics keenly I saw absolutely no effect at all. Nothing. That was rather upsetting but I stopped it before it took all my cash.

Test #2: This test was in two stages. I requested a new audio ad and created two new campaigns. One had a budget of $300 for a week and I removed as many chances for it landing on AM stations as possible (removed talk radio, etc.). The second campaign was another budget of $700 for 3 weeks, also removed AM type stations, and requested it play on fewer stations to maximize my bid. Both overlapped for the first week.

The effect: Nearly worthless! There was an effect, but minimal. I had roughly 30 direct visitors a day increase for the one week the campaigns overlapped. After that, nothing. It pretty much dropped to before-ad levels. That's the "good" thing about Audio, you know they're almost always the direct ones.

Totals for these two campaigns before I stopped the second one was 123 ad plays to 418,000 listeners. Out of that, over the week plus, I got about 230 new visitors at a rather sizable cost. The total comes to right around $3 per visitor not including the cost of creating the ads. The only good thing I can really say about my experience is that you get the best of the best visiting your site. They heard your ad and went out of their way to find out more, instead of just impulsively clicking on a link. That turned into some conversions ($1,000 in sales, but cost that to get them!) and good time-on-site."

and Afmusan says:
"RESULTS: Radio ads are not trackable for me. My site is established and already gets 10k people a day. Radio is more about branding. Hit peoples ears over time and maybe they remember your site when they see it online or it comes to mind when they think of my service. I'm looking at the addresses of my customers and trying to find spikes in those markets where the ads are run. About three days into ad plays in two markets, I did notice a few more customers that are from those markets. I don't know for sure, but they could be from radio. I do think that a customer may need to hear your ads a few times to 1) remember it or 2) start to trust it to check it out. I don't think it would do any good to play one ad one time in 100 markets. I think each target market needs to be blanketed over several weeks.

My site uses Adsense at bottom of pages for those leaving the site. From the first weekday run of ads, I did notice an upturn in page views on the site. Not necessarily customers. Is it from radio or just a coincidence? Is it from a search engine adjustment? Is it from added content? I don't know for sure. I'm not seeing ranking changes or new sections adding up to the increase. I'm seeing more direct entries. But maybe something else happened that I can't detect. My site has sections to it with worldwide subsections and I do notice that the increase is in the USA section in the area of the ad focus. I'm not seeing the same increase in other world areas or topics outside the ad focus. The increase in ad revenue from that day forward is more than I'm spending on the radio ads so I can't complain yet."

There are a few endorsement quotes on Google's radio ad site, but those must be taken with a grain of salt, so I think the jury's still out on this one, unless some of you have more positive data?

Monday, September 24, 2007

David McClure: tellin' it like it is

Every once in a while I meet someone and get to talking about work history. They'll tell me where they worked before and I'll do likewise. Usually when I say I worked at Netscape from 1995-1999 I have to quickly add that I joined a week *after* the IPO so they don't make the mistake of thinking I actually made much money.

The way I do this is by saying that when I joined Netscape, I didn't know the difference between a IPO and a PPO. That was pretty much true at the time; remember, this was summer 1995 and the 'IPO' hadn't exactly become the household word it is today.

Since then I've worked at a bunch of startups, and each time I've started earlier and earlier stage, culminating with being one of the first three employees at Efficient Frontier - which has since gone on to kick pretty much everyone's ass in the SEM space.

You'd think that I'd know a ton about stock options, VC funding, liquidation preferences and cap tables by now, but the reality is I've never actually been the founder. I've avoided all that and instead found work at startups with interesting ideas and interesting teams.

The past 12 years has taught me a lot about all this, though, unfortunately the hard way. Things I've learned:

1) be 'in the know', or be screwed. Unless you are involved in the fundraising process directly, expect to never get all the details you need to correctly value the options you're working for.

2) 95%+ of all Silicon Valley startup employees are spineless jellyfish when it comes to sticking up for themselves in equity negotiations. Over the years as I've increasingly been at the forefront of equity knowledge discovery within the startups I worked at, I've also made the concurrent realization that most around me weren't asking for knowledge that management wasn't giving.

3) Absolute power corrupts, absolutely, and in the absence of employee unions and/or legislation, investors and founders hold 100% of power within startups. If you're a VC or a founder, you might be thinking to yourself "Yeah, that's *exactly* the way it should be you punk", but you'd be wrong to think that way, because which startup is successful is all about


and execution doesn't happen without motivated employees, and motivated employees don't stay motivated employees unless they have a reason to stay motivated employees.

All that said, I have to applaud David McClure for his great find & review of Leo Dirac's presentation on VC term sheets and liquidation preferences. If you want to go from being the Silicon Valley equivalent of a Microserf, check out Leo's presentation as well as Dave's great review.

You'll learn about all those things you vaguely were aware of, like preferred vs common, liquidation preferences and hopefully you'll be better prepared in your next startup to ensure that you know everything you need to know to go into that long hard schlog that is startup life with the right expectations.

Monday, September 17, 2007

Google Disables Axis of Evil Geotargeting

Iranian's love their satellite dishes that connect them to external news sources

Iranian President Ahmadinejad does not...

While trolling Digital Point Forums I learned today that it appears Google as of very recently is no longer allowing people to advertise via AdWords in Iran, North Korea, Sudan and Syria from outside of those countries. What's interesting, though, is that they're not saying anything about this recent change. Google says specifically:

"Thank you for your email. I understand your concern that you are not able to target Iran as one of your targeting countries while setting up a new campaign. This is because Google routinely reviews its policies regarding advertising, including its policies regarding the geographical targeting of ads. Based on our most recent review of these policies and consistent with the Terms and Conditions of our AdWords service, we have decided to no longer offer geographical targeting of ads to Cuba, Iran, North Korea, Sudan and Syria.

We apologise for any inconvenience this may cause. Please keep in mind that this decision only concerns our advertising and in no way affects the search results we deliver. Google offers broad access to content across the web without censoring results."

While a person in Iran can still advertise via AdWords in Iran, an American who wants to, say, expose Iranians to Wafa Sultan's revolutionary message, can no longer do so.

Why is Google doing this, you ask? Wouldn't Google want to support U.S. advertisers' ability to target international markets?

I'm not sure why they've made this decision, but I'm pissed they've taken a great tool out of the hands of people wanting to spread viewpoints into the countries whose people most need to hear them.

Someone from Google, please correct me if I'm off base here, but I think this is BAD, right up there with their decision to censor their search results in China. What kind of message does this send to Ahmadinejad, Bashar and Kim when we prevent our own people from targeting messages to their people???

Priuses are great, caving to foreign despots is not.

Thursday, September 13, 2007

Top Ten SEPR Click Percentages Revealed (Again)

I put the leaking of AOL data right up there with the unearthing of Pompeii, the finding of the Dead Sea Scrolls or the discovery of King Tut. Future historians will have a perfect window into the times with this data.

Many of you may have already seen this particular unearthed discovery, but I haven't so bear with me.

Below is data showing the % of clicks going to positions 1-10 on AOL (a good proxy for GOOG since GOOG supplies both paid and organic listings on AOL).

Position 1: 42.13%
Position 2: 11.90%
Position 3: 8.50%
Position 4: 6.06%
Position 5: 4.92%
Position 6: 4.05%
Position 7: 3.41%
Position 8: 3.01%
Position 9: 2.85%
Position 10: 2.99%

1st page: 89.82%
2nd+ pages: 10.18%
[NOTE: sample size is 4.5M+ total clicks]

The data is from an EarnersForum discussion wherein folks there were looking through the reams of data AOL mistakenly made public back in August 2006.

Tuesday, September 11, 2007

Yahoo mobile search

Friday, September 07, 2007

Omniture offers $65M for Offermatica

Back when I worked at Netscape in New York ('98-'99) I got to know the brothers Roche (Matthew & James) who in 1996 started Fort Point Partners, a web development shop that worked with retailers such as J.Crew, KMart/, Intel, HP and Estee Lauder.

It's rare to see co-founder brothers, and even rarer still to name your startup after a surf spot - 'Fort Point' is a great spot under the Golden Gate Bridge in S.F.

Anyways, James & Matt built a thriving practice around helping retailers building personalization into their ecommerce websites, and F.P.P.'s platform of choice became ATG. Lead forward a few years and I was at Totality, a managed services provider (since acquired by Verizon) that provided outsourced management of large [ATG-based] sites such as American Airlines, Best Buy, Sharper Image and Cabela's.

Totality was actually a company formed from within Fort Point Partners. Michael Carrier, who was one of the early guys (co-founder perhaps?) at FPP kept noticing that FPP had no one at the client to hand over site operations to once they had finished building the site, and so he had the bright idea to build a company to do just that - outsourced site operations so that the site operator could focus on their core functions (marketing, merchandising, etc).

FPP hit hard times after the bubble burst, and hobbled along for a few years until they got the religion around multi-variate A/B testing. Not to toot my own horn (well, yes, actually I am tooting my own horn), but I had a conversation back in mid-2003 with one of the FPP guys wherein I told them how interesting and long-legged I though the multi-variate A/B testing space was. My core thesis? Everyone was just starting to buy lots of paid search traffic, which would eventually lead to an arms race on the part of advertisers to increase their conversion rates in order to remain competitive in search.

Since then FPP renamed itself Offermatica, hired one of the best bag-carrying sales managers I know in Darren Johnson (aka DJ) and has gone out and fairly well dominated the landing page optimization industry here in the U.S., despite Optimost's 2-year head start. This I credit to

1) FPP's having been in business since 1996;
2) the hunger that comes from having invested 10 years of your life into a startup;
3) a huge amount of learning in the area of personalization and via the ATG platform;
4) a stellar sales execution effort, to which Matthew and James Roche contributed, but which DJ spearheaded. Frankly, DJ kicked ass.

Seeing Omniture, the world leader in web analytics, pay $65M for Offermatica is for me a heart-warming story of perseverance, the restorative powers of bursting bubbles, brothers-in-arms and great sales execution.

Congrats guys!

Google Analytics Alternative