A new website is born

Earlier this year, I announced that I was redoing my company's website because it was time. Well, much time has passed, and last week I finally 'flipped the switch'. Yay!!!

Here's a short chronology of the process, and my thoughts:

Select the 'designer' and technology:
I knew from the beginning that I wanted to have fellow blogger Mike Boyink do the site. It's really just a coincidence that he is just one town away here in Michigan--I would have picked him regardless. Why? His approach to websites is very reasonable and aware of the technicalities. No grandiose strategies or pandering to my whims, just a very knowledgeable and capable individual.

Mike's weakness, if you want to call it that, is that he just has one hammer in his toolbox, a blogging/CMS application called ExpressionEngine. He justifies is as just a very robust database creation tool, with the flexibility to do a lot of things, which it is. His expertise in applying EE is his strength, as he literally is the guy who wrote the book on using it.

Laying out the site:
While I knew what I wanted, I had some choices to make about content and categories. Categories drive navigation, and making the categories make sense to the visitor is important, but must also fit the product line. With that high-level thinking done, the rest of the site fell into place easily, relying on what already existed and worked on the current site.

Categories >> Families >> Products

Designing the site:
Mike, like myself, doesn't profess to be a 'designer'. But what he provided was certainly nice enough. The site was designed in Photoshop then sent out to be created using compliant CSS. We agreed to go with flexible-width website.

Building the site:
The database and the design template were done quickly, and that left me to just add content. Product specification data was entered by hand, which was much easier since I had someone else here do that part. But there was a ton more stuff to add.

Things slow down:
For most product lines, copy-and-pasting descriptions was not that simple. Carry-over code had extra tags and also didn't always get rendered by ExpressionEngine the same. And, of-course, in-line links had to be corrected to new URLs. The database structure meant I had to fit the content to the template, a big change from my previous freedom to create a product page however I wanted.

I spent a lot of time working on the product line descriptions. It was time well spent, however.

Learning EE:
In the process of building out my content, I also started exploring ExpressionEngine on my own. I tweaked the page templates a number of times to better suit the way I was working the content, especially the way the product specifications lay out. I started to learn why Mike likes EE.

Fresh pictures:
Another slow-down was getting new pictures. A number of products had changed looks, so I had to format new pictures for the site. And some existing images weren't the right size, or had other problems. Plus, I was adding photo-galleries for each product line, so there were additional images to prepare. I spent a long time working on images that I really hadn't expected.

Finishing it off:
I allowed almost a month to review the product content over and over, as well as build-up the non-product content. I drove myself crazy with continual tweaking, but have no regrets. When I finally flipped-the-switch, it was almost anti-climactic. Its been live a week now. Google has already slurped-up the 301 page redirects in their site results, and none of my pages have lost any ranking.

The secret:
Okay, you guys know that I don't directly mention my employer here, but if you wanted to see the site, and while you are at it, get Boyink's side of things, you would want to visit this post here.

I'd have to say there is nothing spectacular in what we did or the resulting site—it is basic functional brochureware. But it is a lot of work to get everything right and am proud of the results. Thanks to Mike Boyink for all he did—I certainly recommend working with him if you need a site of a similar scope. Now I can start thinking about adding some important tools that can leverage the power of the database we've created.

Haven't we figured out browser sizes yet?

Another rant about website technical issues:

Screen resolution IS NOT the same as browser size!

Years ago we figured out the difference between 'hits' and 'page views', why can't we figure this one out?

Google Analytics can't. Their description of their 'screen resolution' report says:

"Which screen resolutions do your visitors use? Optimizing your site for the appropriate technical capabilities helps make your site more engaging and usable, which can lead to higher conversion rates and more sales."

WTF?

Wait, Jakob Nielsen must have some good detail on browser size:

"Do not design solely for a specific monitor size because screen sizes vary among users. Window size variability is even greater, since users don't always maximize their browsers (especially if they have large screens)."

Okay, we are getting somewhere here. But Jakob's article goes on relying solely on screen size data. Here's a bit more:

"Starting at 1600x1200, users rarely stretch their browser windows to the full screen because few websites work well on such a wide canvas. Big windows are magic for working on spreadsheets, graphic design, and many other tasks, but not for the current paradigm of Web pages. Today, big-screen Web users typically utilize their extra space for multiple windows and parallel browsing. "

Yea, exactly. I have a 1280px wide screen (19" flat), but run my browser at about 950px wide, so Jakob's 1600 number is a bit high. Or am I just not normal? Unfortunately, as long as analytic programs like Google's focus on screen resolution we don't know.

Some digging, and I found someone who cared to do some research on this subject. Thomas Baekdal took the time to actually physically measure (funny video) users browser screens. Here is a great page with bar-graphs of screen sizes and browser sizes. The full report is here. In summation, he says that to fit 95% of users' browsers, you site should work in this size range: 720x400 - 1408x912. And these are actual content window dims, as the full browser window will actually be larger.

Wow, 720 width is a lot smaller than the 1024 width that Jakob Nielsen says you should target (or even the 800 minimum he identifies, which Baekdal's report shows only 85% of users will fit).

And Baekdal's report shows that I am not normal, as only 25% of users with my screen size run their browsers un-maximized. But 25% of any target market should be too large to ignore. So ignore all that data in your analytics and study Baekdal's report instead.

Gizmo: Use the MeasureIt Firefox plug-in to measure objects within your browser.

Sherpa chimes in on online pricing

I spent a significant amount of time last winter discussing the issues of published pricing on your website (1, 2, 3, 4 and more). Now Anne Holland of MarketingSherpa chimes in on the subject: Should You Reveal Pricing Online? Overcoming Fear & Loathing.

She sums up in a few paragraphs the challenge with pricing on your website from the perspective of marketing and avoiding the issues of 'why not', while acknowledging that they exist. To the discussion, she brings this thought worth thinking about:

"Your prospects will find pricing information even without your help. They’ll ask friends at other companies, post queries to industry email discussion groups and boards, ping analysts or surf the Web researching.

The only problem is, you’ve now lost control of your pricing messaging. You can’t surround the conversation with value and branding. You can’t be sure that the correct information is even getting to prospects.

And they’re making those decisions before they agree (or not) to meet with your sales reps. Because pricing information is now sought much higher up in the sales funnel than most marketers suspect."

Wow, she's probably right. If you want to know the price, you are more likely to ask a friend than an supplier first. And she specifically calls out research that says this is true for 50K+ technology/software buys. While to us as vendors, it seems fool-hardy to go with hearsay pricing, that is exactly what she says prospects are doing.

That explains the people who finally call looking to buy with a fixed budget that is half what they actually need. One more reason why pricing info is so important as early in the game as possible. How can you get that info to the prospects?

I hate making labels with Mail-Merge

We have a modest Christmas card list. Mainly sales reps and agents and key customers. For a couple different reasons, I couldn't just print the labels from our CRM program. It should be easy enough to do mail-merge labels in Word, right? Hardly. And there is no way I would expect my secretary to be able to deal with the problems I encountered. Here are the problems and solutions:

• Export from CRM in .dbf format is fine, but it establishes a 'database name' and range of cells. Because I moved around some columns after importing into Excel, those columns were no longer in the range and wouldn't import.
  Solution: In Excel: Go to Insert>>Name>>Define (How arcane is that?)
  

• My mostly-blank 'Address2' column imported as zeros. Weird. A little Googling to find out that Word has decided the column is in 'number' format.
  Solution: In Excel: Select the column, then Format>>Cells>>Text
  

• I have my target's names in a field called 'Contact', but the mail-merge tool doesn't recognize this. It is looking for 'First' and 'Last' name fields. The nice thing is that the rest of my column titles it does recognize and inserts them all correctly.
  Solution: Rename the column to 'First'. (You can also manually map the 'First' field to the 'Contact' column.)
  

• Whoa! When I 'preview' the results, everything is double-spaced. I can only guess that the label template is also influenced by Word's Normal.dot file.
  Solution: Ctrl-A to select-all, then change the paragraph settings. (At this point, I also changed the font size to 9-point to make room for long labels of overseas addresses.)
  

• Then the labels print way to close to the top of the label, and some actually get cut off. While this gives some extra room for those overseas addresses, it is no help if half their name is cut off.
  Solution: I go to 'Page Setup' and change the top margin from 0.5 to 0.6
  

Wait, there is one more.

• Labels look all good, but I am only getting one sheet to print. And to add insult to injury, when I print, the printer ignores that I have selected 'manual feed'.
  Solution: Turns out I hadn't finished the mail-merge. Despite having what looks like good-to-go labels, they are actually just a 'preview' of the data. The last screen of the wizard gives you a choice to 'print-out' or 'edit individual labels'. You have to choose one of these to complete the merge. Of course as soon as you select 'print-out' you get a printing dialog box and no way to know how many sheets of labels you need. Better to go with the 'edit' choice.
  

So there is my Word mail-merge label troubleshooting guide. Now that I have written this, maybe I can have my secretary do it next time, if I give her a copy.

Can I quote you on that? “ No â€

People are getting sloppy about their quotations online. No I don't mean bad reporting, I mean bad typography. Smart quotes gone awry.

I've seen on websites, blogs, and newsletters, an increasing amount of weird characters. Characters where quotation marks or other punctuation should be.

The culprit seems to be cutting-and-pasting from Word or other Office programs. Word creates 'smart quotes' that automatically adjust to whether they are on the right or left of a word to form the proper look. I've had a love/hate relationship with these for years, as it looks professional until you are using the smart-quote to indicate feet or inches, which should look just like they key on your keyboard shows.

What I didn't realize is that Word is substituting its own font-characters for the smart quotes, which isn't web-standard. So we end up with a 'lost in translation' problem, like sending the same text repeatedly thru a translation tool.

Examples of how quote marks can render:

• Bad: cut-n-paste from a webpage with smart quotes: “Noâ€
  
• Ugly: Type directly into Blogger: "No"
• Beautiful: Hand-coded into Blogger: “No”

Now, I'm not exactly sure of what programs do what to quote-marks. I'm sure it differs as to where you are copying from and where you are pasting to. Other places you might have to hand-code the characters. Be aware, and adjust accordingly. In my case, My new website's CMS seems to render from hash-marks to smart-quotes automatically, but pasted smart-quotes get ugly.

Here are a couple articles on the subject:

The trouble with Em 'n En @ A List Apart
Smart Quotes @ Wikipedia

Clipped off!

UPDATE (1/7/08): I just got a response from the first service I sent files for 'free sample clipping paths'. Um, that's over a month for what was promised as overnight service. Why did they even bother?

I like to do things myself. Where it is easy and I have the tools. But I've found that outlining images in Photoshop isn't as easy as it looks. It takes time. It can take quite a bit of clicking and dragging.

I'll just send it out, I think. Must be folks out there doing clipping/masking online. Yes, there is.

I looked at a couple sites in the search results and they proudly declare they are using folks from India or other places to do this for just a couple bucks per image. Sounds like a good business model, and they are honest about it. Overnight service too, based on time differences. I am ready to commit.

"Try our service for free" the site declared. So I did. Submitted two images and a note saying I had more I'd like done. That was a week-and-a-half ago.

Guess what happened? Nothing!

I went to find another company earlier of this week. Look, another offer for a free trial. So I submitted the same images. Same result--nothing!

What do I do? I obviously don't want to work with these unreliable companies. Do I just keep running down the search results? Seems like almost as much work as doing it myself.

My next step? I'm going to try the company advertising in AdWords for this service. They are at least committed enough to spend money to get business.

UPDATE: The AdWords company had the image clipped in less than an hour for free. They are in the USA, too (and make a point about it). They even added a drop shadow like I asked for. I think I found the winner.

Eight random things

Swell, I give Larry a plug for his podcast, and he returns the favor with a 'tag' for a blog-meme thingy. ;-) Seems, I am up to share eight random things about myself. I'm sticking with recent things.

The Paperwork

When you are tagged by me, you must link back to this website. Then when writing your meme, post the rules before your list, then list eight random things about yourself. At the end of the post, you must tag and link to eight other people.

Now … let the fun begin!

The Meme

1. Current song stuck in my (and my family's head) Ding Fries are Done (sorry)
2. Item I really want on my Christmas list: Genius Graphic Tablet (poor man's Wacom)
3. How I spend my Saturdays: Making powerpoint slides for my church's worship service. Lots of fun surfing Flickr for pictures.
4. Recently celebrated 18 years of marriage on 11/18 by finding our #1118 geocache. At night, after the kids went to bed. Nerds. (Click for a pic and story.)
   
5. Just had to order a new iTrip FM modulator for my iPod-mini, via eBay. The current one started going all crackily, so must have a loose connector somewhere.
6. My wife finishes her Masters of Library/Information Science degree in two weeks. The light at the end of the tunnel. Anyone need an archivist?
   
7. Right now, listening to a Randy Newman Cover Story via Coverville podcast. The Tom Jones track at the end is very cool.
8. My cell-phone is almost four years old. Its an antique. Still holds a charge, and drops don't hurt it. I look it as two bonus years of not having to step foot in a cell phone store.
   

Tags
I'll tag some quality marketing bloggers worth noting for this silly little game:

Rick Hendershot
Mike Gerholdt
Pat McGraw
Paul Gibler
Gary Slack

Oh well, that's five. Close enough.