Jump to content


Discussing Web Design & Marketing Since 1998


Hall Of Fame
  • Content count

  • Joined

  • Last visited

  • Days Won


projectphp last won the day on October 24 2012

projectphp had the most liked content!

Community Reputation

5 Neutral

About projectphp

  • Rank
    Honored One Who Served Moderator Alumni

Contact Methods

  1. https://www.cre8asiteforums.com/aboutcre8asiteforums-page/ > Industry experts Ammon Johns (aka Black Knight), Jill Whalen, Bill Slawsk joined as Co-Administrators and assisted with the re-launch. All three left to pursue their own careers and professions, with Jill even launching her own High Rankings Forums. Missing an i.
  2. projectphp

    2016 By The Ad Business

    Hola. Late to the topic but I wonder if people worry about this too much? If 50% of clicks are fraudulent, then the value of a click drops 50% if people track ROAS. If no one tracks ROAS, then that is more the problem than fraud.
  3. Ammon's back - excellent! How have you been? I'm going to go in a different direction. I think there are two extremes that you can go to with any marketing, including SEO: 1. Those that try to sell ice to Eskimos. 2. Those that would rather move to a desert and sell ice there. I have complete respect for the "sell ice to Eskimos" crowd, but to me, that seems like a more difficult task than it needs to be. It pretty much flies in the face of the origin of the word marketing - which is from the noun market which is a place you go to sell where people come to buy. Why would you try to sell things people do not want, in a place they don't want it? And why is that a GOOD idea? Why not simply fulfill a need and, more specifically, create new and fresh demand? the best SEO is whomever is responsible for this: That is ~400 million searches a month, all of which these two businesses rank not only first for, but dominate the first page. And that is just the one word keyword - there is probably close to a billion searches with amazon and ebay as part of the keyword phrase SEO works best when branded search is high. Branded search converts the best, delivers the most consistent results and is the least likely to be at risk of algo changes. Branding is the holy grail of all marketing. That's what real marketing is, and that's what great SEOs help achieve - better search coverage for businesses to help grow their brand, so next time someone searches for [Amazon] instead of [books]. None of that is possible, though, if the user experience of buying / interacting is poor, and the short term SEO that ignores the bigger business issues is putting a bandaid on a severed leg.
  4. projectphp

    Panda & Penguin Recoveries - A How To Guide

    > what are the possible causes? The big thing a few people (such as Jim Boykins) were trumpeting at Pubcon was that (and I just got off a 14 hour flight and feel like I look so this is not entirely accurate) that one of Penguin and Panda (I forget which) was about links, and the other was about user signals like bounce rate. One interesting comment was about pogosticking - which is where users go SERP -> click listing 1 -> click back to the SERP -> click listing 2 -> SERP -> listing 3 etc etc. When users don't find the answer they want, they keep looking. The theory was - and no one I have as a client has suffered much so this is pure hearsay - was that if you don't provide users with what they want for a specific query, then you start to drop in position. As a really good example, if the first result for "do blind people dream" (https://www.google.c...nd+people+dream) leads people to come back to Google and click the second ranked page, that's a pretty strong signal that the 1st ranked page isn't solving the user's query. FULL DISCLOSURE: I have no idea what will work, and I truly don't think anyone else does either, so I'd really be cautions about what you do to solve issues. That said, one exercise with minimal possible downside that you can do is check keywords with a high bounce rate in Analytics, and see if the page people get to actually solves that query. If not, look to beef the page up to improve the page's ability to serve the user for that query, e.g. if they get to a page with cheap flowers, make sure you have your CHEAPEST flowers on that page - not try to upsell. That is so very low risk an activity that I have no problem recommending, in fact it is a good idea fullstop. Whether it actually helps solve your black animal issues or not, I truly can't vouch for.
  5. Maybe it is a spammer? They often use big sites as a referrer to mask who they are. Is this ogfiles of Google Analytics or some other JS based analytics too)?
  6. The issue is likely spaces. You could try this to check: $date1 = date('Y-m-d', strtotime(trim($product["departuredate"])) ); That removes whitespace left and right. Also, always store dates in a DB as Y-m-d, ALWAYS. If you have 10-05-2010, you get year ten. Change from date to datetime, you'll be glad you did. Nothing else need change, and you'll be able to do day to day searches superfast.
  7. Never use strings, instead use timestamps (or int) or datetime. Then you can do: $days = 2 * 24 * 60 * 60; // two days in seconds $start_date = time() - $days;$end_date = time() + $days;$sql = "SELECT * FROM table WHERE departureDate >= {$start_date} AND departureDate <= {$end_date}"; Timestamps, even if you just use midnight, work infinitely betterer, are smaller and, as an int, easier for SQL to work with (number comparison is an order of magnitude faster than strings).
  8. Depends what bounce rate means. An SE could ONLY measure it as: 1. Keyword specific. 2. As it relates to a RETURN to the SERP. Anything else would be too difficult to do. Thing is, IMHO it is seeming more and more likely that there are now multiple algorithms, on separate searches, and bounce rate for extremely high volume searches is probably a really useful metric, at least IMHO. Do they use it? No idea, but I would wager they will one day, but only on high volume terms.
  9. IMHO, there are two main elements to Social Media Marketing: 1. Social Media Optimisation (SMO) - A friend of mine put me straight on this: SMO is about making sure your site / a section of the site is ready for Social Media. think RSS feed links, digg reddit links etc. It also includes making sure you have a page on sites that allow it - A FaceBook page, a MySpace page, a LinkedIn profile, and that you own the Twitter handle for your business name etc. This is a first step, and isn't about perfection, or doing a great job, it is about making sure you own the real estate, and have a presence when and if you decide to do more. Most businesses do a pretty average job at this, I must say, and it really is about a one or two day process, including research time, to cover 80% of the biggest sites. 2. Social Media Engagement - this is actually using the various options to market, engaging in the conversation, responding to events etc etc, and is much more involved. Some businesses will benefit greatly from this, others not much at all. The business case for SM is simple: people talk about you whether you want them to or not. You can choose to engage them, and do so on terms you can at least influence, or let the conversation go on without you. the degree to which you participate, and need to participate, is or should be dictated by your other business goals and requirements. The low hanging fruit of SMO is something most every business should do, but the more involved, complicated and time consuming second stage is something most businesses will not get a heck of a lot from, to be honest. To put some specifics behind this, Dell does a great job and likely benefits from it greatly, but what % of their budget does SMM get? My guess would be a fraction of other areas. A local restaurant, OTOH, should monitor reviews, have the pages I listed (FaceBook etc) and maybe, and only maybe, offer lunch/dinner specials via twitter each day. That's about it. In between, a medium sized retail chain selling mountain bikes, which won't have a lot of one off specials, would benefit from SMO, and perhaps some sort of interactive help section on their site, posting on a blog and being part of a local forum. Thing is, the time it takes to do the latter may not be that beneficial, compared to say doing taxes, and in the near future (say the next 2-3 years), might keep the idea on the back burner, as a "nice to do one day, but not resourced today" idea. Each business is unique, and whilst the benefit of social media is real and increasing all the time, the ROI, beyond basic SMO, will usually lag behind other activities and needs to be understood in that context.
  10. Wrong question, because you get into the whole pointless "whitehat" definition debate. Better question: what are the consequences and do I feel comfortable doing it? Personally: Google respect my robots.txt, I'm happy to respect theirs
  11. projectphp

    Backup For Peace Of Mind

    I am really p***ed that web devs haven't made version control easier to use. When we get XML documents, life will be a lot easier. I love Git, because it is really simple and efficient, and with a central repository, you can grab copies on multiple computers. That is pretty much teh ultimate in portability, and anyone doing web wok that DOESN'T use a VC system (Subversion, Git, heck even CVS) is nuts.
  12. The irony is that the shining SM test case is Dell and Twitter. Ironic, because the value is measured precisely (Dell sold XYZ million in ABC monthsthrough Twitter - figure of months and revenue varies). I think the reverse is the issue: no one knows when NOT to SM. The list of "should do SM" is really every business minus the "should nots". And I have never heard anyone argue when NOT to do SM, which to me is vital in understanding the value proposition. There was an article I read today via Hacker News: Freemium did not work for Phanfare that had an excellent list of when Freemium makes sense: To me, that is the sort of starting point SM needs: when to do it, when not. IMHO, the starting point of measurement is first deciding if SM makes sense at all for a specific business, and then looking t how it makes sense, and last measuring that "sense". Examples: Dell & Twitter: track sales. "SEO Guru": Track re-tweets and followers, blog mentions, links, Sphinn counts. Clothing Brand: measure re-tweets and followers, blog mentions, links, Sphinn counts and have direct response ads (coupons etc), "soft" metrics old skool style. Local Plumber: Don't bother. Plenty to potentially measure, and perhaps even equate back to some bottom line
  13. http://www.techcrunch.com/2009/07/11/real-...ten-social-crm/ timely article. I am not sure you can measure SM in anyway other than traditional offline metrics, and even then, it comes down to a vague relation to any bottom line or revenue impact. I think that is why I fimnd it so frustrating and, oddly, why so many traditional ad and pr agencies find it so wonderfully. Finally, they have an online marketing chanel they "get", measured best using techniques they are used to working with and, imprtantly, can charge for in addition to the work.
  14. Monopoly is a hard thing to definbe. Airports are the worst. The price of food in an airport is ridiculous, and you are stuck there!
  15. That's the problem ultimately. It is hard to measure truly effective marketing. Word Of Mouth is so flippin' random! The problem with this is that it is hard to divert monies to areas where you can't be sure of a return. Social Media is trying really hard to not fall into this trap, but unfortunately, it has so far chosen rather dodgy methodology to justify itself. Which is a shame, because there is value there somewhere, I'm jut not sure anyone knows precisely how to quantify it.