Disqus - Latest Comments for jonathanmendez

Re: AdTech Expo Hall – Big Business, Big Spending, Little Innovation

scotthoffman — Fri, 06 Nov 2009 13:12:24 -0000

Jon was great to see you as well. Thanks for the comment...it is very appropriate. "ad:tech best represents the dark underbelly of the web marketing"

Re: AdTech Expo Hall – Big Business, Big Spending, Little Innovation

jonathanmendez — Fri, 06 Nov 2009 13:01:35 -0000

Thanks for the mention Scott and great bumping into you there!

Indeed I think you sum it up when you say that it has become a more a trade show and less a conference. I would say that might have even been a larger dichotomy in that regard at ad:tech SF (which in a strange way doesn't feel that long ago).

For me ad:tech best represents the dark underbelly of the web marketing. But maybe what we've learned this week from the Zynga and Facebook flap is that the dark underbelly may be the only belly where there is fat.

Re: What if online business model innovation is slowing down?

jonathanmendez — Mon, 28 Sep 2009 10:07:41 -0000

IMO real innovation is about creating digital channels. Email, Affiliate, Display, Search, Mobile, Social. Every advertiser wants new channels. Some channels have built native models and others have borrowed. Of course the native models kick ass and the borrowed models are weak in comparison. I hope that Twitter can take their $50M and build a native model different from anything we've seen.

There's another huge issue here and that's the digital acumen of people holding the bulk of the media dollars. In short they haven't a digital clue and the kind of advertising they do does not lend itself well to a user controlled medium. I'm not sure that gap will be bridged without radical innovation on both ends. A tall order.

Re: Online advertising is all about purchasing intent

jonathanmendez — Sun, 27 Sep 2009 17:11:31 -0000

you are correct and the comparison does make your point (and of course I agree 100% with the premise here) but I think it's important to note that disambiguating intent gets much more complex - even with the query "cameras" it's likely the smallest intent segment are people that are ready to purchase. For advertisers there is also keyword match types that must be factored.

here's some previous elaboration specifically about targeting recovery v discovery http://bit.ly/XOpxM

Re: Online advertising is all about purchasing intent

cdixon — Sun, 27 Sep 2009 10:26:42 -0000

good points. but you do know that when someone types "cameras" they are a lot more likely to have purchasing intent than when they type "abraham lincoln".

I'd be interested to hear you elaborate more on the discovery vs recovery concept.

Re: Online advertising is all about purchasing intent

jonathanmendez — Sun, 27 Sep 2009 10:24:30 -0000

important to keep in mind that it's impossible disambiguate intent from most keyword queries. i would say the reason search works is because google is able to blend the SERP with both discovery and recovery (intent) content in a way that delivers user defined relevance - making everyone happy all the time. really it's only after the click that you can truly understand purchasing intent (and thus the failing of most online advertising is after the click).

Re: How to keep visual design consistent while A/B testing like crazy

jonathanmendez — Mon, 21 Sep 2009 17:06:44 -0000

great topic andrew. thought i'd chime in based on some A/B testing experience.

you can use radical differentiation as an A/B starting point to get the "central design vision" but once you have that vision tested and validated then move to a more iterative elemental approach. this is best served by a wireframe to scale with defined element areas. it doesn't mean elements need to be there nor should it define what or how many elements are present but it lays a visual pixel foundation that can be built upon and tested against. once that is accomplished you can then roadmap out the testing ideas and form a strategy in advance than can be iterated against.

your amazon example brings up an important distinction between the two modes of optimization - presentation of content and delivery of content. having already determined the optimal layout amazon is now testing content delivery and your point is spot on that this needs to come first in the testing cycle because visual cues such as layout are a much stronger factor of influence on performance that the content itself.

i've done a bunch of flow testing and offermatica had (and omniture t&t has) a "steps" progression defined for most every test. usually this is rich in actionable data and follow-on test ideas. for a brief time we had a "hosted flow" offering where we would architect and a/b test multiple flows. we had some amazing results for blockbuster video doing this. it remains a sweet opportunity but there are huge technical hurdles to overcome and it only makes cost sense with high volume/value transactions. still, some vc should fund this pronto.

Re: The NY Startup Scene

fredwilson — Wed, 02 Sep 2009 16:02:06 -0000

Yes. Same is true of finance

Re: The NY Startup Scene

jonathanmendez — Tue, 01 Sep 2009 12:19:27 -0000

NYC is (and will continue to be) the epicenter of advertising & publishing. As these mediums experience sea change it is inevitable that homegrown start-ups will be the ones that help revive the old and create the new.

Re: What We Can Learn From Mess

hymanroth — Tue, 25 Aug 2009 15:10:55 -0000

Good points

Re: What We Can Learn From Mess

jonathanmendez — Tue, 25 Aug 2009 08:57:28 -0000

Thank god Craigslist made it through the 2.0 era without adding a social network on top of it as many would have suggested. What always strikes me about Craigslist is their email proxy - such a simple idea yet who else has leveraged this as well? Never underestimate the value of the underlined text link.

Re: To my first 10,000 blog subscribers: Thank you!

jonathanmendez — Fri, 21 Aug 2009 18:30:34 -0000

Congrats AC! We started blogging about the same time. I'm not sure people really understand how monumental getting 10k subs is. With a comparative lowly 3k I do. :) Keep it going brother, the web needs your POV for its survival.

Re: Agency Demand Platforms

jonathanmendez — Wed, 15 Jul 2009 22:50:57 -0000

Thanks Darren. This is a much needed POV.

I would only add that there are incredibly difficult run-time challenges as you begin to patch the platform ecosystem components together and create massively parallel distributed systems. The challenges increase exponentially if your platform is optimizing in realtime (under 250ms).

Re: The Difference Between Total Uses and Active Users

ShanaC — Wed, 15 Jul 2009 16:14:33 -0000

Actually I find it more complicated it that- you can't trust what people are saying. You have to go, sit down, and watch them, and then ask about those behaviors, and then try some more watching.

If you just did focus groups and asked- there would be no ATMs. People apparently hated the idea when you asked them. You need to sit down and watch people.

Further, a lot of design should stem from the body-the hardware is at least an extension of what your hands, mouth, ect are doing. Software, being in some the most apparent extension of your hardware, should be really easily understood as being somehow, in some sort of abstract or real way, to the body and its systems, and the contexts it is put in. I always get shocked/unshocked by how much and how little thought there is put into software based on how humans work in context to the situations they are and the bodies we proudy have.

Re: The Difference Between Total Uses and Active Users

fredwilson — Wed, 15 Jul 2009 14:31:19 -0000

Interesting point. How can you know what people have not done? User testing, focus groups, surveys?

Re: The Difference Between Total Uses and Active Users

fredwilson — Wed, 15 Jul 2009 14:29:26 -0000

Interesting point. How can you know what people have not done? User testing, focus groups, surveys?

Re: The Difference Between Total Uses and Active Users

jonathanmendez — Wed, 15 Jul 2009 09:24:36 -0000

Not sure I agree. I think this sentiment is a shortcoming of analytics based approaches to site optimization. Analytics tell you only what people have done. Often times what's more important is knowing why people haven't done something. This can only be attained through direct observation & communication with your audience (& I don't mean surveys, rather talking and more importantly listening). Inside those answers are frequently unknown issues, missing features & untapped strategies for conversion.

Re: Referrer Processing for Content Sites

jonathanmendez — Wed, 08 Jul 2009 09:57:31 -0000

Yes, the more modernly managed sites are becoming dominant but they are also more distributed (mobile, third party) which presents a host of other challenges. The opportunity for source based LPO is growing and taken on a whole might even be bigger than keyword. What I do know is that targeting to source has a major positive impact on performance.

I don't disagree with your second "objection" either however my point is that it is not the technology holding back referrer processing for content but skillsets and mindsets.

I'll disagree with your last point. I think the depth of data, types of assets and amount of processing required - at least for larger content sites - full integration with their content (in order to reclassify it) is necessary if a fully dynamic experience is being created. Can you do pieces of it outside the domain (articles, ads)? Certainly you can and that's fine but it's not the level of self-learning, intelligence and relevance delivery that's possible, even now.

Re: Referrer Processing for Content Sites

jonathanmendez — Mon, 06 Jul 2009 21:10:48 -0000

Ah -- an issue dear to my heart. I'll answer the question. Never.

Dynamic LPO is still not very prevalent in e-commerce and where it is done it's usually some javascript calls that change a few default content boxes. When done effectively it's usually in a hosted environment away from the headaches of IT which means publishers are not moving on this until more dynamic internal CMS solutions are in place (Mark Logic comes to mind). This gets back to the "never" answer b/c I only see this working as a ground up pure CMS play. Semantic solutions offer a glimmer of hope but most are still years away from where they need to be and scalability is a giant issue.

Technical issues aside how many content sites understand how to really optimize for marketing/revenue generation? This is something they've left to third-parties far too long and is now biting them on the ass. Do they even care about the "next click?" 25% of NYT.com traffic exits to Google yet all NYT bitches and moans about is the 25% that arrives from Google.

I'm really disappointed you dropped the kw feature but I beat on the content LPO drum for years and never saw smoke so it's not surprising. Here's a presentation I gave three years ago at OMMA that fell on deaf ears http://bit.ly/27S9e3 . Most content people don't understand the value in delivering great experiences through relevance. At the end of the day that is not what their business is about. Their mindset is a broadcast and the technology that delivers on that promise will never mesh with a web where the real value is being created simultaneously for the audience, publisher and advertiser in realtime.

Re: Aggregate, Curate, Publish To Create Local Media

fredwilson — Fri, 26 Jun 2009 17:53:21 -0000

Patch is nice but I don't see how it scales

Re: Aggregate, Curate, Publish To Create Local Media

jonathanmendez — Fri, 26 Jun 2009 15:07:20 -0000

thank mark. i just backed your numbers out to eCPM. In any case if you guys can do RPM in the $5-$15 dollar range your clearly on to something big and congrats. As you probably know large pubs like Yahoo have RPM in the pennies and more targeted vertical pubs (say forbes.com) are in the .25-.75 range. Maybe local (and by extension mobile) are the home run.

Re: Aggregate, Curate, Publish To Create Local Media

markjosephson — Fri, 26 Jun 2009 14:00:08 -0000

Hi there. Small clarification: the model is actually based on RPM -- Revenue per 1,000 Pageviews. There should be mulitple impressions per page. In our experience, we're seeing that is attainable target. Today a lot of that inventory is going to national advertisers looking for local audiences. There is a lot of innovation happening on the "solving local advertiser dilema" front and we're starting to see some interesting traction. See Clickable, AdReady, Yodle, etc....

Re: Aggregate, Curate, Publish To Create Local Media

jonathanmendez — Fri, 26 Jun 2009 08:52:42 -0000

Fred- The P&L model assumes $7 eCPM. That's an incredible assumption especially for the types of advertisers that would be relevant to this content. Also local ads are typically "intent" driven services stuff or classifieds, both that lend themselves to a different non impressions based models/opportunities/issues. Have you seen what Patch is doing in this space? Really like their sites.

Re: Directory of Blogs by Entrepreneurs

fnazeeri — Tue, 02 Jun 2009 10:43:33 -0000

Jonathan, thx for posting. Great blog. I added you to the list (and the bundle) at #23.

Re: Directory of Blogs by Entrepreneurs

jonathanmendez — Tue, 02 Jun 2009 09:12:43 -0000

Hopeful you'll consider my blog for your directory
http://www.optimizeandprophesize.com/

Cheers,

Jonathan