12 Hacks for Picking Web Apps

The world of web services has mushroomed in the past few years. This is great for buyers, except the vendor selection process has become more time-consuming.

You can easily build a comparison chart in a spreadsheet with feature comparisons, integrations, support and prices, but you’d miss the qualitative part of the story.

[Read more...]

Six Questions to Ask Your Web Developer About Technology

My morning’s reading started with a brief discussion on Reddit about the relative capabilities of two website content management systems, WordPress and Drupal.

There is no definite conclusion to that battle, but the prevailing sentiment is that WordPress is easier to learn and cheaper to build, and Drupal is more powerful. Excuse this oversimplification, since my topic today is a bit higher-level.

If you’re a business owner intending to build a new website, what should you first ask potential web developers about the choice of system that would power your website?

I’m only listing questions about the choice of technology package — not about the web developers’ business track record, design chops, cost, and familiarity with comparable clients.

The obvious first question, of course, is “Can the CMS do what we need it to do, now?”

Asking these additional questions about the technology will give you a better picture of the total lifecycle cost of your website, and the tradeoffs associated with each.

Six Questions

1. How extensibile is it? After you launch the new website with a given feature set, how easy will it be to add features later on? Examples would be discussion groups, a non-English language, user commenting, a blog, more social media integration, and photo slideshows. Most popular CMS’s have plugin code modules for these features which make future upgrades easy. But ask your developer to investigate how their choice of system will support of your future needs. This discussion would be most productive if you have a rough 24-month roadmap for your site.

2. How well does it support mobile users? Your website needs to look good on iPhones and iPads and Android devices. Does the CMS have a good reputation for supporting this kind of alternative display?

3. How much trouble will it be to migrate my data out of the system? Website CMS’s have a lifespan. What was cutting edge five years ago (e.g. Movable Type, PHPNuke, ColdFusion) is now functionally obsolete. At some point you will need to rebuild your site using a new system, and the portability of the data on your web pages has a nontrivial impact on the upgrade cost. The good news is that most of today’s popular CMS’s store website content in a database, which fundamentally supports data portability. “Flat” websites built by the likes of Dreamweaver are not as futureproof.

4. Will my marketing manager find it easy to make changes to the website? A well-designed editorial interface reduces the risk of mistakes, and reduces the incidence of support calls to your developer. Make a list of the content that you might need to modify on the website before asking this question. Note that if you intend to outsource all your website changes to the developer (or his delegate) then the editorial UI shouldn’t impact your decision much.

5. How big is the development community for a given CMS? Put another way, how many people are available to work on a given system? Your web developer may eventually disappoint you or get out of the business altogether, leaving you with a temporarily unmaintained system. If you have a popular CMS, then it’s more likely that you will find a solid replacement person to work on it. Success is a virtuous circle for software. WordPress is the leader on this score, since it built a broad developer base during its early years as a solid tool for basic blog sites. See this Google Trends chart for one measure of the relative popularity of CMS’s.

6. Does this CMS have a good reputation for security? No web developer will admit to offering an insecure CMS which allows your website to be compromised, but there is a question of degree. Generally speaking, the inherent security risk of a CMS decreases with its developer base. Linus’s Law says that, “given enough eyeballs, all bugs are shallow.” (Another benefit of success.) On the other hand, the most popular CMS’s are bigger targets and thus more inviting to the hacker community.

Bonus Issues

These topics are less critical to the choice of technology, but perhaps worth covering with your developer.

Cleanliness of code. The issue here is whether your CMS builds its pages with tidy code to save bandwidth and display quickly to the user. Drupal, as much as I like it, can produce giant hairballs of HTML & CSS on each page unless the developer is rigorous when building templates. On the other hand, throwing a little extra money per month at the problem (as I mentioned here) can alleviate the problem. I should add that rendering pages for mobile devices is an exception: you want those pages to be lightweight. A good modern CMS should be able to build mobile-ready pages with a more limited set of markup.

Search Engine Optimization. While SEO is obviously important for most websites, the differences between the major CMS’s in their SEO-friendliness are either marginal or overshadowed by factors like your domain name, the quality of your content, the number of inbound links, and site speed.

Accessibility for viewers with disabilities. This in practice usually means legally blind people with screen-readers. See here and here for more.

Don’t Get In the Weeds

Addressing all these issues will help you make a better choice of back-end system for your website, which reduces your total cost of ownership and makes customer acquisition more efficient. These technical questions may also sharpen your online marketing plan itself.

Yet the technology choice shouldn’t outweigh the business issues, as pointed out by the fellows at this web agency. Choose a web developer mostly based on trust, track record and cost.

Why have a website when you have Facebook?

cake2My English friend Frank makes cakes.

What begun as a hobby is now a day job. So he made blue polka-dotted business cards, and over coffee on a Sunday morning he handed one to me.

Under his email address was a URL. Was it cakesbyfrank.com? Nope.


Did he have any intention of building a “regular” website?


Your Company, On the Web

Companies that need an online presence (meaning, um, all of them) can either:

  1. Build a standalone website.
  2. Build a standalone website AND make a Facebook page.
  3. Make a Facebook page.

Most people are at #1, and moving to #2.

Frank made me consider, is choice #3 viable? Can you run a business with just a Facebook page?

Since I’m a consultant, the answer to hard questions is nearly always the frustrating but honest “it depends.”

Yet we so enjoy breaking down the Pros and Cons, and it’s a good thought experiment.

Evaluating the Facebook-only Approach

Customers. It always has to start here. If his customers are already on Facebook, then he’s not shutting business out. (One has to sign up for Facebook to see the delicious confections.) Frank targets upmarket people who have a big social event, so there is a sociographic match there. Is a Facebook-only web approach making it harder for the customer to buy? No, it’s the same as if he had a regular website. Verdict: No Difference Here.

Frank's Cakes_1299624644400Social Proof
. Buyers want to see market validation. On regular websites, this takes the form of testimonials, client logos, awards, twitter followers and so on. On Facebook, the social proof is right there, with pictures of “friends” running along the left column. This is dynamic and believable, and a major plus for the Facebook-only business site. Verdict: Easier and more personal on Facebook.

Startup Cost. Custom websites cost in the thousands (your mileage may vary), and you ordinarily need to spend a nontrivial amount of time with the development team. DIY websites (e.g. WordPress, Weebly) can be free or have a nominal cost, and they take about the same amount of effort. Verdict: Facebook is free, but so are other options. Tie.

Maintenance Cost. Facebook shines here, since the system is so templated and they have hundreds of developers working on the interface. Frank has uploaded photo galleries of his cakes without worrying about FTP and HTML. You can make basic updates to a Facebook page from a phone, or from any computer. WordPress has maintenance easy. Verdict: Facebook in a narrow win.

Intangibles/Positioning: The entire Facebook ecosystem is built around 1:1 personal relationships, obviously. So using FB as your site makes the most sense if your business is in that mental space. Like career coaches, personal trainers, churches, PR people, and event planners. In these jobs, you’re selling yourself as much as the service, and Facebook works best when there is a personal voice behind the business message.

Here, however, we arrive at the first big elephant in the room with the Facebook-only approach. Doing so risks giving the perception that your company is too cheap to have their own website, are technically challenged, or are just too new. The best way to counteract this is to have a lushly designed, popular Facebook page with several engaging apps. Attaining this requires either talent or money, which means you’ve lost some of the cost advantage. Verdict: It Depends.

Lead Generation: Interestingly, Facebook may have an advantage here. Advertising within Facebook is obviously possible — with its excellent targeting options — and you can run AdWords on Google with the FB page as the landing page. So you get to advertise use the biggest PPC search network, and the biggest social media network, all with one “site”. As for SEO, with a Facebook page you probably give up some upside on very competitive terms. For narrow “long-tail” terms like Princeton NJ Wedding Cakes, however, a Facebook page could theoretically break into the top three. (Disclaimer: I am looking for more data on this.) Verdict: It’s a wash.

Conversion and Measurement: At this writing, Facebook pages have no ecommerce component, so you’d need to complete a sale offline or at a third-party website. Facebook’s Page statistics are improving but come up short next to the commercial stats platforms. Plus, measuring conversions is much easier when you have your own site. (That said, if optimizing page conversion is important to your bottom line, it’s time for your own site.) Verdict: Facebook loses here.

facebook-logo1Branding and Control: 
A company Facebook page is always in the Facebook wrapper, which dilutes your corporate image. We can expect Facebook to give Page editors greater visual control in the future, but that will naturally bump up against their desire to keep some consistency of experience. Similarly, even though Facebook’s Pages can accomodate more widgets and functional applications, you can always do more on your own site. Verdict: Better to have your own website.

Security: Another elephant. If your Facebook account is compromised, all sorts of mayhem can ensue. Recovering access is time-consuming and uncertain. True, there are single points of failure in the website world as well (your registrar account, your email account), but the authentication safeguards are better, and you usually have recourse to a staffed help desk at your website host. Verdict: Facebook loses here. Use complex passwords everywhere.

Copyright and Ownership: Last elephant. Facebook owns the content on its site, not you. Even though their usage policies are not likely to affect you in the short term, this fact carries obvious business risk of the “unknown unknowns” variety.

What About Frank?

The choice is not really whether to build a standalone website or a Facebook page; most companies eventually grow into needing both.

For entrepreneurs like Frank, the real question is, “which is better to begin with?” The best candidates for the Facebook-only approach are small, young, local, businesses whose sales rely on a strong personal connection. Other than that, small companies are better off beginning with a WordPress site.


iPhone and Drupal

iPhone integration is coming fitfully to the Drupal world. We are happy to report that two tools, one recently released, are making it easier to create content and maintain your sites. BlogPress is a fairly well-established app that permits you to write simple posts and upload pictures. Exempli Gratia:

You are mostly limited by your tolerance for the tedium of thumb typing. Drupad is a new app & module combination that eases some common site moderation tasks, such as user permissioning, comment handling and backing up.

The above was composed entirely on the phone via Blogpress. Line breaks are absent, and it’s not feasible to include links. This really limits the utility of the app. Yet the picture integration is pretty good: the app grabs a pic from your phone and sends it off to Flickr. So, it’s worth keeping in the hope that a new version fixes that problem and adds features like CCK integration.

I’m quite pleased with the early state of Drupad. Handling comments is great, and that’s the kind of idle drudgery you’re likely to do on a phone while on the train, in line, in the car etc.

A nifty utility


Copying and pasting text between applications can be troublesome (at least in a Windows environment), because copying text also copies the formatting rules along with it.

So, if I copy something from a website and paste it into Word, the text is often in a mismatching font, color, size etc. Word allows you to “Paste Special | Unformatted” but that is several clicks or mouseactions.

Usually I want just the text, formatted to match the destination.

Enter PureText, a tiny program that allows you to use a hotkey  (default is WINDOWS+V) to paste text without formatting.

I think I use it once an hour. Go get it.

Nine Marketing Tasks You Shouldn’t Be Doing Anymore

For the marketer at a medium-sized company, the development of web-based apps has made some tasks obsolete. With some effort and startup costs, by 2009 you can stop:

Running your business without an intranet. A simple intranet can help you share files with remote staff, have discussions, host a wiki, and remain very secure. Open source website software Drupal is a perfect tool for this. Its user and group permissioning is particularly easy, and you can easily extend it into an “extranet” where your partners and commissioned sales reps get access to a portion of the site’s content.

Guessing at what your customers are saying in public. A few ideas: 1) Do a Google search for “better than [your-product-name-here]” or “similar to [name] — remember the quotes. There are a number of variations to this for tire-kicking customers; i.e. “also considered [name]” “looked at [name]“. 2) Yahoo Answers, especially if you cast your net to include competitor names. 3) an RSS feed of your keywords from Tweetscan. 4) an RSS feed of a Bloglines search. The signal-to-noise ratio of these channels can range from bad to good depending on your company name and sector, so it will probably require some tuning. And you’ll need someone to draw generalizations from the little data points. (Hello, summer intern!)

Paying to host webinars. For the medium-sized business that only does the occasional webinar and remote presentation, WebEx is a luxury. It’s a great service, but you can get the same effect with Zoho.

Asking someone else to change some text on your website. In 2001, websites were usually made on a PC and uploaded to the webserver. Dreamweaver was the norm, along with tools like FrontPage. Altering a page (e.g. adding news, updating a headshot) meant asking the IT guy or the web developer, who might have been offsite. Thus began the dance of request prioritization and email tag for reviewing the page. Now, there are sophisticated and near-free content management systems (CMS) that live on the webserver. To make a change, you log into the system through a web browser, and make the change yourself. The key is that you (and the other editors) can be permissioned to make only certain low-impact changes, and changes can be rolled back if you screw up. Good CMS’s include my current fave Drupal, Expression Engine, Joomla, and Movable Type.

Relying on a PR firm to monitor the news for you. A PR person still helps interpret news and coaches you through a response, but there’s no reason to lean on them as the source of breaking news about you and your company. For media monitoring, Google Alerts and Factiva ($) will get you 90% covered.

Paying lots for development of embeddable website widgets. Seen all over Facebook, widgets might make sense for your marketing. If your widget audience is in the tenthousands, then it probably makes sense to get professional agency work. But if you’re just running an experiment, DIY widgetmaker Sprout is worth a go.

Wondering which half of your advertising works. I’m looking at you, Ogilvy. There are a bunch of ways to reduce that “wondering” to a third or a quarter. Namely: Landing pages, conversion tools for PPC ads, custom URL’s for print ads and DM campaigns, and the filtering tools in web analytics. Integrating lead management tools like Eloqua and Salesforce helps you track conversions.

Speculating on how people get to your site. Knowing where your web visitors is not a new development, but it’s worth mentioning here for the smaller companies whose website statistics are accessible only through their webmaster or web developer. The point is that you can have online, free access to reports that show you the following: 1) which other sites are linking to you and how many people are coming from them; 2) what locations your visitors are coming from; 3) which ads they clicked; and 3) what search keywords your visitors used. Google Analytics is the best free tool out now, but there is a new free package called Piwik that bears watching. With GA, You can add fiters and goals to track these people through to conversion.

Paying for shopping cart software. I admit that the integration and setup of a cart will cost you much more than the cart itself. The bigger point is that quality of free ecommerce software is a lot better now than it was several years ago, and better-supported by a developer/user community. Old and confusing: OSCommerce, ZenCart. New hotness: Magento. Our pals at Avenue Verve, who are our preferred developers of ecommerce sites, swear by Magento.

Web hosting highs and lows

UPDATE November 2010: MediaTemple’s Grid system continued to be very flaky in 2008 and 2009. Downtime and a security problem drove me away, into the capable arms of <a “href=”http://www.wiredtree.com”>WiredTree.

The majority of the websites I maintain are hosted at MediaTemple in LA. They came recommended by Komra at Design4Results, answer the phone, and are big enough to give me a secure feeling. In September they had some problems with their clustered web hosting service, called the Grid service. Some latency, duplicate emails, and the resolution process lingered on a bit. The Grid has never seemed rock-solid, and so they’re on a shorter leash with me.

But I write here not to bury MediaTemple, but praise them. During the problems they sent out updates consistently, via trouble ticket, blog thread, and RSS. Good for them. Hiding such obvious problems from well-networked buyers — who are skilled searchers — is asking for trouble. And after the system got back to normal, they sent out a credit via email.

As per the last incident update on (mt) Media Temple’s incident tracker system, 24 hours have passed and we have seen no further latency issues on (gs) Cluster.2
(mt) Media Temple is issuing 1 month’s credit to customers affected by this issue (INC# 285). We would like to take this opportunity to once again apologize for the unexpected access problems. We understand that our customers run web dependent businesses on our systems and that slow or inaccessible websites or email are simply unacceptable. We would also like to convey once more that this incident has spawned numerous internal reviews, new monitoring points and new adjustments to our cluster growth formulas. The primary aim for (mt) Media Temple is for this to never happen again.
We thank you for patience and your continued business. Regards, Demian P. Sellfors CEO (mt) Media Temple, Inc.

The blog post reads a bit differently. Including the CEO’s name in the email had a cooling effect.

Moving from the sublime to the ridiculous, I present the final communication from Web Wide Media. They are a Texas-based budget host I inherited for one of my pro bono projects. In the highly probable event their service announcement disappears with their business, here is the relevant bit.

Ahoy! Like a modern day pirate, I am assuming command of an abandoned ship….
Let me introduce myself. My name is [name redacted by Andrew]. I own and operate DoorCountyHosting.com, and previously DoorHost.net until it was sold in April of 2006.
I’ve been working with WWM.net as a support tech, and a client, for almost a year. In recent events, I’ve had contact with the owners of WWM.net only on three occassions since June of 2007. One of those times was a two way communication. The others were one sided quick notes from [name redacted by Andrew] , and nothing further.
I’m assuming WWM.net to be an abandoned company, and taking control of it in an effort to ensure the clients with the hosting services and support they have paid for. Why is it assumed abandoned? As you can see – the cPanel licenses have gone unpaid, as well as the servers. Neither myself, nor the data center have a reliable method of communication to the owners. Have any of you heard from anyone but me since June??
However, there are many complications with this take over.
First, I’ve told [name redacted by Andrew] in an email, just moments ago, that I will give back the company if/when he wishes within the next 30 days. I am not out to hurt him, his income, or his company. If 30 days pass with no contact, the company will continue to be assumed abandoned. And any attempts to regain control will be denied without legal order to do so.
Second, I need to work out a time frame with the Data Center on how long I can keep these servers online before payment must be made. I will assume no responsibility for payments due before todays date. Whether that will be acceptable to the Data Center is unknown at this time.
Third, DO NOT EXPECT MIRACLES. I will need time to assess how many clients are left, and how to best organize the clients on the servers for minimal expenses and optimal server performance for the clients’ web sites. This means that you should likely expect some of you to have IP changes and custom DNS name servers to need updating. Feel free to post questions and comments in the general support forum. I will answer/update as soon as humanly possible.
Fourth, I’ve assumed control of the billing system. As of 12:20am CST (-6 UTC) all credit card processing, and PayPal payments are being handled by me. Any payments made before this time on 10/26/2007 are out of my control and refunds are impossible. There are a select few that appear to have already been set into a batch for processing, that I can not stop or access. So billing date for those will show up as after this date/time. If a refund is requested, and you fall into that category, you will be notified at time of refund request.
Lastly, THIS IS NOT SET IN STONE. Again, I need to consult with the Data Center and/or [name redacted by Andrew] may show up in the next 30 days. Think of this more as a letter of intent….
That is all for now. Expect things to turn around a get better in the near future.
Cordially, [name redacted by Andrew] WWM.net

The ensuing thread is a guilty pleasure — but only because moving my site from WWM won’t be a problem.

How SEO can screw up your…

Some Search Engine Optimization (SEO) tactics can give you a marginal boost in rankings, but hurt you on a net basis.

The main culprit is usually an overweening preoccupation with on-page keyword optimization, but link-building can hurt too.

This article is not about black-hat SEO practices, which are covered in plenty of detail elsewhere. Instead, we’ll look at “good” SEO practices when they’re overdone.

Let’s see what gets hurt.

Muddied Product Positioning
Pretend you sell software to corporate accountants. Corporate accounting software is obviously your #1 target phrase, but you naturally care about capturing related searches. So your SEO advisor says, “hey, let’s target some other key phrases too.” Next thing you know, they’ve added pages on your site titled Corporate Controller Software and Shareholder Accounting Softwareand Corporate Asset Management Software, in part since those are phrases that the keyword suggestion tools returned. And then the SEO advisor includes links to those pages within your left-hand navigation, because that helps the new pages rank higher. Your problem reveals itself when a prospect visits your site, and sees all those links. Prospect then thinks, “well, this company’s product does a whole lot of different things!” And that may not be the market positioning you want.

Underperforming Website Copy
There is an allegedly optimal level of “keyword density” that makes a given page rank well. For example, if you have “organic tea” occurring six times on the page, that’s better than two occurrences. Visitors read differently than search engine spiders, however, and they still respond to effective copy. Stuffing pages with key phrases, then, may reduce the conversion power of a page. So your tradeoff is higher traffic versus page effectiveness. Would you rather have a page that gets 100 natural search referrals and one conversion, or 30 referrals and two conversions? It’s possible to have both, but how are you going to get there? It’s much easier to test and change copy (and design) for improved conversion than to test keyword optimization. We think there is more upside with conversion-building efforts, and other SEO tactics. Landing pages for pay-per-click ads are a great place to test copy and design.

Cluttered Website Navigation
The left nav, footer, and site map page get heavy weight with search algorithms, since they are treated as pointers to a site’s most important pages. SEO advisors often will load the navigation with links to search-engine-friendly pages. This distracts people from your intended sales funnel. Excessive intrasite linking means people go in circles, and get frustrated. Repetition of keywords and phrases in links will lead to guesswork by the prospect: “Is the thing I want behind this link or that one?” Every site visitor has a finite number of clicks they’re willing to spend at your site – don’t make them click more than they have to.

Noise in Measurement
Your SEO scorecard should be visually simple, and focussing on the biggest business drivers. Don’t let it get cluttered with ranks of long-tail search terms in second-tier engines. Doing so obscures the high-value Google terms, which – like it or not – is where the money is. Track infrequent terms on a second worksheet if you must, and present an aggregate performance metric for the lot of them on the scorecard. Also, don’t track search terms which bring unqualified traffic.

Confusing Site Maintenance
This is a minor factor that can be avoided, but we’ll mention it anyway. Regular content changes to web pages is a healthy thing for SEO. For sites not built with a content management system, one byproduct of changing content is an accumulation of defunct and unlinked files on your webserver; the “retired” versions of live pages. You might even have new and old style sheets. Over time this can clutter up your webspace, to the point where it costs your webmaster time. You should have a system for saving old files (in case you need to revert back to them, or audit for troubleshooting purposes) that is clean and organized. Moving them to a hidden directory is one way. Giving them a unique extension like .defunct is another, since sorting by type gives the webmaster a method of visually ignoring the retired files.

Repetitive Blog Entry Titles
If you’ve got a business blog, resist the temptation to consistently pack your entry titles with keywords. Include keywords in the basenames instead, and save the keyphrases for your marquee blog content.

Meta Descriptions
Inside the top of your web page, there is usually a field for a description of the page. It looks like this:
<meta name=”description” content=”The leading purveyor of premium tea”>
Google often uses the copy after content= as the blurb under your listing in a search engine results page. This is a marketing opportunity, a chance to help position your product effectively before the click. Enthusiastic SEO advisors may instead see the META description field as a chance to further pack your page with more keywords. Yet doing so may mean a less-well-performing blurb, which means poorer conversion from the search results page.

Blackened Wikipedia Reputation
Even though a link from Wikipedia to your site hasn’t counted in the search algorithms since 2005, it’s still a potential source of inbound traffic. Don’t go adding links to your site within Wikipedia content, however, unless your content has the right patina of independence. Wikipedia editors are quite mindful of commercially motivated edits, and squash them with a Puritan zeal.