Steve Smith's Blog

Hiring Inbound Marketing Score CARD

ssmith — Tue, 09 Feb 2010 17:16:00 GMT

In their book, Inbound Marketing (review), authors Shah and Halligan use a couple of acronyms that, maybe due to my military background, I thought could be improved. The first one was VEPA, which I thought made a lot more sense as PAVE. PAVE relates to qualities of a call to action, and it is easy to remember that if your call to action has these qualities it will PAVE the way to better results.

In chapter 12, the authors describe a “framework…for hiring and developing inbound marketing savvy employees.” Like VEPA, this four-letter acronym is a nonsense word, DARC. Again, it’s easy to see how this can be improved by converting it into an actual word, CARD, and then leveraging a common use of this word to really make the mnemonic stick by making it a score CARD or report CARD.

The authors already make the point several times in the book that it’s important and fairly easy to measure online attributes of web sites and key company personnel, so it only makes sense that the CARD qualities be measured as well. The qualities are:

Content Creators. At the heart of Inbound Marketing is the need to create remarkable content. You need people who can do this.

Analytical Chops. Almost everything related to online marketing is measurable, to the point where sifting through the data is a challenge, You need people who are skilled at analyzing data and finding out what’s working and what isn’t, without resorting to guesswork.

Reach on the Web. It takes time to develop a following online. Wherever possible, you want to favor candidates who have an extensive following online, particularly as it relates to your industry. You can easily track your key employees’ and prospective marketing hires’ LinkedIn connections, Twitter followers, and Blog Subscribers.

Digital Citizens. Hire marketing people who are fluent with the web, not merely familiar with a few phrases. If they’re not into RSS, twitter, blogging, and whatever the latest new thing is this year, rank them below other candidates who clearly live online.

So, use the CARD acronym to remember how to score potential inbound marketing hires. Who knows, maybe I can be involved in their next book as acronymsmith or something.

http://www.flickr.com/photos/pjern/ / CC BY-NC-SA 2.0

PAVE the Way to Effective Calls to Action

ssmith — Mon, 08 Feb 2010 16:53:00 GMT

In their book, Inbound Marketing (review), authors Shah and Halligan describe some key traits of effective calls to action. The four important qualities of killer calls-to-action are that they be Valuable, Easy-to-Use, Prominent, and Action-Oriented (chapter 8). The authors go on to suggest that this be referred to as “VEPA”, but I think we can do better than that. By simply shifting the letters by 2 positions, this becomes the much more memorable mnemonic, PAVE:

Prominent. Your offer must stand out. It should be difficult to miss. Don’t make potential customers have to think about how to interact with you.

Action-Oriented. You want the visitor to DO something. The call to action should compel the user to take action (hence the name) using a verb. A great analysis of the difference wording can make can be found here.

Valuable. Make sure it’s clear to the visitor what’s in it for them if they take action. Internet users have (justifiably) grown skeptical of wasting their time with offers online. Make sure your offer if valuable but also instills trust and helps you build a relationship with the user.

Easy-to-Use. Make it as simple as possible for the user to respond to your call to action. Think Amazon One-Click purchasing. Again, don’t make the user think about how to proceed, and don’t put any unnecessary barriers (like longer-than-necessary registration forms) in front of them.

So there you have it – four qualities of calls to action that will help PAVE the way to higher response rates for your web site. Perhaps when it comes time for a second edition of Inbound Marketing, they’ll use the PAVE acronym (and maybe even credit me).

http://www.flickr.com/photos/myconstructionphotos/ / CC BY-NC-ND 2.0

Inbound Marketing and Small Business Trends

ssmith — Thu, 04 Feb 2010 16:30:52 GMT

I recently read Dharmesh’s and Brian Halligan’s Inbound Marketing book, which has a lot of good tips for startup companies to follow in order to maximize their online reach and popularity. Many of the tips are pretty obvious: start a blog, get people linking to you, build a following on twitter. But there are some that are pretty easy to overlook, and at the end of the book is a very good appendix that is a checklist for any startup or small business to follow.

One of the key points that describes the importance of Inbound Marketing, is that it is difficult for your competitors to one-up. First, let me explain that inbound marketing differs from outbound in that you are helping customers who may be looking for you or a service like yours, rather than reaching out to potential customers and interrupting them during whatever they’re doing and trying to drag them to your site. There was a great quote from Dharmesh that actually prompted me to buy this book:

With Pay-Per-Click…You’re basically at the mercy of the stupidest market entrant.

If the only reason people are coming to your web site is because you spend a lot on advertising, and in particular, on PPC advertising, you have no barrier to entry protecting you from competitors. In fact, the more efficiently you tweak your PPC keywords to ensure you’re getting a great ROI out of them, the more vulnerable you are to a stupid new competitor who has no idea what these keywords should be worth, but who has a pile of venture capital cash to spend building initial market share. When this happens, the effect will be instantaneous, and without warning.

With inbound marketing, on the other hand, over time you build compelling content and spread the word about this content via social networks like your own blog, twitter, youtube, facebook, etc. Over time, you amass a large amount of such content, as well as a large following of customers and fans who make up your community. If a new player enters the market, whether they spend a bunch of money on ads or a bunch of effort building up their own inbound marketing efforts, you will see them coming. The effects will not happen overnight, and any advances they make on your market share will be gradual and hard-fought.

I had a chance to hear Anita Campbell speak last night in Hudson, Ohio, where the agile consulting company I founded has its offices. Her presentation was on “2010 Small Business Trends and Opportunities” and included a lot of good information that related to the idea of inbound marketing, I thought. She also recommend a couple of other books I’ll probably pick up: The Social Media Bible and The New Community Rules.

Anita also talked about a few different sites that I want to check out (so this is sort of my own notes-to-self to follow up on):

PointBanner.com – custom banner design, simply and easily.
Knowem.com – search over 350 social media sites for your company/brand/user name to see if it is available. For a fee, they’ll register/signup the username on the sites for you.
GetListed.org – Check local listings for your business, add it to local listings. Great for businesses with a physical address.
SBA – Government Contracting How-To – Site with courses on getting started on government contracts
Register in Government Contractor Database – What it sounds like.
Retargeter – ads on popular sites that redirect your own users back to your site, for a flat monthly fee.

While I’m listing resources, another couple that are good from the Inbound Marketing book are:

Website Grader – Generates an SEO report for your web site with optional comparison to competitors.
Twitter Grader – Same idea, but for twitter.

The book recommends, and I agree, tracking key individuals’ and your company’s results on a somewhat regular basis, so you can see whether you are improving. You should track things like your twitter followers, inbound links, total indexed pages on your web site, etc. and chart them over time (perhaps relative to your competitors) so you know if your social/inbound marketing efforts are succeeding with some concrete results. Here’s an example twitter grader result:

Another trend for small businesses is the use of cloud and third-party services to support the business. For instance, it’s very popular now to get customer feedback via a third party service like UserVoice or GetSatisfaction. Need to track customers and leads? Many businesses love to hate SalesForce.com. Need a blog? There are plenty to choose from. Help desk? Check out ZenDesk.com. Customer surveys? Try SurveyMonkey.com. Planning for an event? Try EventBrite, which I use to organize the monthly Hudson Software Craftsmanship meetings. Need a newsletter? Try ConstantContact. Whereas a few years ago businesses would invest a lot of time into building these features themselves, today you can easily outsource these commodity functions to third party sites that focus exclusively on this problem, leaving you to focus on your core business.

Azure Tip: How To Deploy a ZIP File to Windows Azure

ssmith — Wed, 03 Feb 2010 21:27:00 GMT

Last month, The Code Project ran an Azure contest and gave away several Amazon Kindles. As part of the contest, which we hosted on Azure, we deployed a sample project with all of the necessary install files for getting started with Windows Azure. It turned out to be slightly more difficult than expected to actually get the zip file deployed to the cloud, so I thought I’d post here in case others ran into the same issue.

The issue isn’t related to ZIP files, of course, but to any content file that is added to a web project that isn’t a typical ASP.NET or web file type (e.g. .aspx, .gif, etc.). For instance, if you simply add a .zip file to an ASP.NET web project (in this case, an Azure Web Instance project), it will have project properties like the box to the right.

As you might expect, this won’t actually be deployed. At first glance, it might seem like the right thing to do (especially if you’re in a hurry and are trying things quickly to get this to work) is to change “Do not copy” since obviously we do want this file copied up to our Azure deployment. And you would be wrong.

Simply put, the issue here is that a Build Action of “None” means the build and deploy features of Visual Studio will treat your file like a social pariah and ignore it completely. Pretend your file has leprosy – Visual Studio wants nothing to do with it. But that’s not how Visual Studio treats “cool” files, like images and CSS, which it also doesn’t need to build. Looking at these for reference, we see they look like this:

So the fix to get your ZIP files and other “content” to be deployed to the cloud as part of your Azure deployment package is to simply specify a Build Action of “Content.”

Incidentally, this month (Feb 2010) the Code Project is resuming their contest, this time with an even bigger prize: an HP TouchSmart TX2-1370US 12.1-Inch laptop computer. All you need to do to enter is create a new Azure application and register it with the Code Project’s site (using the template we provide makes this easy). Check out the Code Project Feb 2010 Azure contest details page for more info.

Prevent Resharper From Adding Regions

ssmith — Wed, 03 Feb 2010 20:32:10 GMT

A couple of days ago I was annoyed that Resharper was insisting on turning my abstract base NUnit test class with nothing in it but a shared [SetUp] method into a one line class with a collapsed Setup / Teardown region in it. While I didn’t always feel this way, my experience has taught me that regions are a smell in your code. They are a way to hide things you don’t want to deal with or look at. It’s kind of like putting makeup over a melanoma instead of having a doctor remove it. Here’s a pretty good analysis of why regions are a code smell if you’re interested.

So anyway, I’m using Resharper and am in general a huge fan, and I know I can configure this thing every which way, but I look through the options and nothing jumps out at me about how to adjust Region settings for NUnit tests. I manually removed the region and checked in my code and, rather than spending more time and effort researching a solution, griped about it on twitter today. Not long after, Todd Ropog let me know about this:

Prevent Resharper from adding Regions to Interfaces

Chris outlines the steps nicely:

Go to ReSharper | Options

In the left explorer pane, find Language | C# | Formatting Style | Type Members Layout

Uncheck "Use Default Patterns"

A huge nasty XML document appears. As Kyle mentions, Don't Panic.

This is why I never found what I was looking for – if you don’t uncheck the “Use Default Patterns” checkbox, you never even see the XML used. Once you see this:

you’re pretty much there. All you need to do is look for things that say:

<Group>
  <Name Region="Some Region Name"/>
</Group>

and remove them. You can remove the whole <Entry> if you like, but if it’s doing other things like <Sort> and you want it to continue doing that, you’re better off just removing the <Group> or <Group>’s contents.

So, to sum up:

Regions can indicate a code smell and should be used sparingly.
Resharper is a fabulous tool…

…but it’s so darned flexible that finding how to do some things can be hard.

Twitter can be a great way to get other people to tell you how to do things you don’t know how to do. :)
Hopefully this helps a few people Googling/Binging or having their friends Tweet how to remove regions from Resharper code cleanup in the future.

How to set up TRIM with Win7 and SSD Drive

ssmith — Mon, 01 Feb 2010 15:40:25 GMT

I have an Intel X-25M SSD in my developer workstation machine (and it’s quite fast). However, I’ve heard from others that over time SSD performance can degrade due to sub-block level fragmentation that occurs as a result of write combining. Fortunately, newer SSD drives (like mine) support the TRIM command, but of course this only works if your system is sending the command to the drive.

How Do I Know if Windows is Using TRIM for my SSD?

Fortunately (and thanks to Ken), there is a simple command you can execute to determine if windows is sending TRIM commands to your SSD drive.

Run the following command:

fsutil behavior query disabledeletenotify

If the result is 0, then Windows is sending TRIM commands to the device (which obviously needs to have firmware that supports TRIM for this to matter).

Party with Palermo: MVP Summit Edition

ssmith — Fri, 29 Jan 2010 00:28:27 GMT

The Code Project is sponsoring the next Party with Palermo in a few weeks at the MVP Summit. If you haven't already, sign up. The cost is a nominal $5, just to try and keep the RSVPs accurate. You can see who else is coming on the sign-up page (via EventBrite), so check it out. Hope to see you there!

Product Idea: Polarizing Plate Covers

ssmith — Sun, 24 Jan 2010 03:52:00 GMT

I’m one of those people that is always coming up with crazy business or product ideas. The problem is always that there just aren’t enough resources to go after every idea, and I at least know that

An Idea is Not a Business

so at least I don’t pretend that maybe some day I’m going to capitalize on these things. So on the way to CodeMash last week talking to Brendan about Mythbusters, apparently the show did an episode about trying to defeat traffic cams that take a picture of your license plate if you’re speeding. I haven’t seen the show, but the end result was that most of the things people have tried (or businesses sell) to avoid these cameras simply don’t work. The ones that do work are illegal, because they’re too opaque or reflective and therefore they violate the laws regarding license plate covers. Seems like you’re screwed (unless you have a really really fast car).

But the thing about these plate cover solutions is that they’re all static – if you put hair spray or glitter or whatever on your license plate, it’s just there, all the time. But if you put a little LCD plate over top of it that was totally transparent, but responded to a bit of electricity and became more opaque as more electricity was applied, you’d quickly get to the point where, at some speeds, your plates were simply impossible to read.

Thus, if you simply were to put a magnet on one of the car’s tires (like any bike computer/speedometer), you could probably generate enough current without any batteries or external power source to make the LCD plate cover opaque, gradually, at higher speeds. If you get pulled over, of course, the plate is completely visible, and at lower speeds it would be completely or mostly visible as well.

Update: Turns out this is such a good idea others have already invented it – there are several on YouTube.

Windows Azure Pricing and Shared Hosting

ssmith — Fri, 22 Jan 2010 13:59:00 GMT

Windows Azure, Microsoft’s cloud computing platform, has recently gone into production and will begin charging customers next month. You can keep up on Azure news and blogs at AzureFeeds.com, a community moderated resource. One of the promises of Azure is to treat application hosting like a utility service, through which one pays for what one uses, just as with electricity or telephone usage. In fact, you’ll find that, like your phone plan, there are many options to consider when trying to estimate what an application will cost while running on Azure, but there are also many ways to test out an application for free (such as via MSDN Premium). You can learn more about Azure pricing here. There’s also a pricing calculator application for Azure, written in Silverlight and hosted in Azure, that you may find useful.

It’s a Hotel, not Electricity

Currently Windows Azure costs $0.12 / hour for “Compute.” There are other costs for Storage, Storage Transactions, and Data Transfers. The latter three are all pretty self-explanatory, I think, but Compute has been a source of some confusion. Many assume that this corresponds to some kind of CPU load, and that therefore if they have a low-impact application that sits idle (but on and available) most of the month, they would only need to pay for a small fraction of the actual time their application was running. However, Windows Azure Compute time is calculated like a Hotel room – you’re paying for the reservation whether you use the room or not. You will pay the same for your blog that never uses more than 2% of the CPU as you will for a busy data-crunching application that’s constantly at 80+% CPU. Just like you’ll pay the same for a night booked at a hotel whether you end up crashing there or just staying out all night. In both cases, you’re consuming resources by virtue of the fact that they are reserved for your use, and cannot be committed to other customers.

Minimum Monthly Cost

If you’re thinking about hosting your blog on Windows Azure, you should understand that it’s going to probably cost more than your ultra-cheap shared hosting plan at one of the many .NET hosting companies. But that’s not really an apples-to-apples comparison, since Windows Azure is all about having an abstracted application fabric that allows for linear, elastic scalability. That’s not available for $9.95 from your hosting company’s shared web site plan. Last summer, Chris Pietschmann calculated that his blog would cost about $99/mo to host on Azure. I’m not sure that’s still exactly correct now, but I think it’s still in the ballpark.

The Hotel Metaphor Breaks Down

If you’ve ever travelled with a bunch of buddies on a road trip, conference, vacation, or whatever on the cheap where you’ve all shared one or two hotel rooms, you know that your per-person cost on a hotel can go down dramatically when you start packing people into the rooms. That $99 hotel room split 5 ways is under 20 bucks. Likewise, if you decide to host your blog, along with several other family members’ or coworkers’ blogs, and maybe a couple of other hobby apps you’re working on, all in one Azure instance, now that relatively high price tag starts to get very reasonable. Unfortunately, the current version of Azure requires everybody to have their own hotel room. Your web Azure web role is going to have one, and only one, domain address (e.g. http://codeproject.cloudapp.net), and that corresponds to one ASP.NET application.

Now, you’re certainly welcome to write ASP.NET code to detect what URL the user entered and render something different accordingly. However, that’s a lot of work and tends to make it difficult to use third party applications (like an open source ASP.NET blog framework, for instance). What’s really needed is a way to have a single Web Role in Azure allow for multiple different web applications, just as a single virtual server can host multiple web sites (each with a separate IP address, port, and/or host header). I hope that we will see support for what I call a Shared Web Role in v2 (or 1.1?) of Windows Azure.

http://www.flickr.com/photos/lori_greig/ / CC BY-ND 2.0

Axosoft OnTime and Queues

ssmith — Thu, 21 Jan 2010 14:24:17 GMT

At CodeProject we’ve recently adopted Axosoft OnTime for our task, feature, and bug tracking needs. We had tried a number of different solutions, and there was even some discussion of building our own (naturally), but in the end we’ve settled on OnTime, at least for the time being.

OnTime breaks up items into Defects, Tasks, and Features by default, and you can establish fairly rich workflow rules for each of these that tie into actions (for example, there is a status for ‘Complete’ but also a workflow of ‘Complete’. If you choose to use the workflow feature, you can make it so that when the item moves into the Complete state, various other things occur, like re-assigning it back to the original requestor so they can Close the item. If you just update the status, you’re just updating a field and nothing else happens).

OnTime supports many different projects and sub-projects, and your view into the system is typically filtered by where you are in the hierarchy. So if you go to Project A which has subprojects 1, 2, and 3 you will see everything related to A, 1, 2, and 3. If you move to subproject 1, you will only see its items. For managers (and developers oftentimes) who move between many different projects, this can make it easy to limit the scope of what’s being displayed. And of course there are many customizations you can make in terms of filters to control exactly what is displayed and how in any given view.

I’m using the windows desktop client as well as the VS2008 plug-in. There’s also a web-based front-end and apparently an iPhone app, though I personally haven’t yet broken down and bought an iPhone (mainly because I like Verizon). The performance has generally been acceptable but not what I would call stellar. I definitely could see room for improvement here.

Queues in OnTime

At NimblePros, we used a variety of systems for tracking user stories, and the one we still use because of its ease-of-use and simplicity is Zen. Zen is ultra-simple and acts as a virtual kanban board. It does this task very well. One thing it forces you to do is to order all tasks in a queue. There’s no notion of priorities (although you could use tagging or color coding for such if you really wanted to) – instead the queue defines the order. This is invaluable and saves so much time.

Yesterday at The Code Project we had a status meeting and discussed a variety of items that were in OnTime with various priorities. At the time, I had about 4 High priority items, one Immediate priority item that I was In Progress on, and maybe 10 Medium priority items, some of which had been there since last month. In going through the meeting, a bunch of these Medium priority items were seen as needing to just get done, so their priority was increased to High. Of course, none of my High priority items went away, and my Immediate priority item remained in its place at the head of the queue. So what was the net effect? Now “everything is top priority.” Of course I made it clear that this was the case and we did in fact establish an ordering of the items, but it wasn’t something we could do in OnTime.

In Zen, if I want to change the sequence of a set of stories, I just drag the story card to its new location in the queue. It’s very simple. In OnTime, I have to try and come up with some kind of priority and/or voting or rating combination such that via some sorting routing the item happens to fall into the order that I want it to be in. What I’d really like to be able to do is replace Priority in OnTime with a Queue order. Then within a project I would be able to arrange all of the items in a Queue (and ideally this Queue would span Defects/Features/Tasks, or at least could do so). Then, in my rollup view of all of my action items (across many projects), I could arrange my tasks in any order I want. I would also want to be able to do this via drag-and-drop.

I mentioned some of these ideas a while ago on twitter (I’m @ardalis), and @hamids and @axosoft were quick to respond, so I know they’re aware of this request. I don’t know if it’s something they’ve heard from other customers or not, but I do know they are adding support for a Kanban view that might just do what I want, so I’m eager to see the beta in February. I’ll be sure to write more about it once I’ve had a chance to play with it.