Truths And Myths About Open Source eCommerce Solutions: What The eBusiness Leader Needs To Know Now

by Brian Walker

Open source eCommerce solutions are increasingly on the radar of eBusiness professionals interested in, and in need of, eCommerce platform technology. There are many drivers behind this interest — from the obvious ones of lower costs, to the desire for independence from vendors and increased resource flexibility. While in the past open source eCommerce solutions were often seen as appropriate for very small businesses, I know many industry observers will agree that in the last year, the buzz around open source eCommerce solutions like those from Magento, Apache OFBiz, and OXID eShop has really picked up steam.

We here at Forrester have looked into the realities of implementing these solutions in the context of a larger small-to-medium (SMB) or lower mid-market range eCommerce business to see how well open source eCommerce platforms are living up to their billing. We identified five core views on open source solutions that are commonly held by eBusiness professionals (which we have dubbed “promises”):

Promise No. 1: Open source software is less costly.
Promise No. 2: The same important core eCommerce features that I need are there.
Promise No. 3: I have more flexibility with open source.
Promise No. 4: I can leverage a community to drive innovation.
Promise No. 5: It will be easier to access resources and do so at a lower cost.

So what did we find after talking with many retailers, developers, and solution vendors? Here are two of those ‘promises’ explored:

Promise No. 1: Open source software is less costly

One of the fundamental attractions of open source software (OSS) is the presumed avoidance of expensive licensing, maintenance, or revenue share deals with vendors. Typically, this upfront and recurring capital expense is the focus when comparing open source and commercial applications. Often the professional services and maintenance costs are secondary.
Truth: You will avoid recurring and scaling software licensing and maintenance fees. Once the product is up and running and the base requirements are met, the costs will be in incremental professional services on an as-needed basis. Licensing, maintenance, or revenue share fees found in commercial products are eliminated. This is particularly compelling for the less mature business expecting to grow into its solution over time.

"The license cost is free. The software stack is a traditional AMP stack — so that was free. The box that it's running on is $5,000. It's not a beefy box, and it supports it just fine." (Project manager, small online retailer)

Myth: I can expect a lower total cost of ownership (TCO). The costs of open source projects are going to be compared with those of commercial applications, where the licensing and maintenance costs are identified upfront and there are predictable professional services costs to launch and integrate. The costs of the open source solution will of course lie in professional services fees or in the allocation of internal resources. Many find that the integration and custom development costs exceed expectations.

"The biggest component is capitalized cost of development. We wish we could get a quicker ramp-up than we did. Two good people took four weeks longer to really ramp up on it, and we've since done refactoring twice in 10 months." (Chief technology officer, small online retailer)
"From a return on investment (ROI) and TCO perspective, I honestly think it is a wash by the time you pay for the custom development needed to get you where you need to be. And you have the downside of the custom development. [The product] is too young. This may change in the future." (Project manager, Web design and development agency)
"The real downside is that it is expensive to host. I thought my cost would come down — it was, in fact, pretty comparable or even more than we would have spent on [a commercial product]. It's such a resource-intensive package; I would love to spend less on hosting costs." (eCommerce manager, small online retailer)

Promise No. 2: The same important core eCommerce features that I need are there.

A quick scan of sites using open source eCommerce platforms and a review of the available online information can lead one to expect that the features present in the leading open source solutions are the same as those found in many commercial applications.
Truth: Open source can provide an effective foundation, enabling focus on differentiation. Although open source eCommerce applications are immature, compared with many commercial applications, OSS can enable businesses to focus on differentiating features over time by providing them with core and commodity needs — such as basic promotions, site templates, as well as basic search, browse, and cart features. This differentiation has led to business results and to great customer experiences, even though this flexibility and extendibility can come at a cost (as can be seen in the aforementioned higher TCO).

"Having worked with a lot of platforms, I believe that it's not going to be right for everybody. You expect to pay for advanced functionality, and [it] has been able to get that base functionality out there for developers to work with. I'm very results-oriented — it's all that I really need. Our next challenges are driving more traffic and raising conversion rates." (eCommerce manager, small online retailer)
"We have been playing with it. Lately we have been splicing [it] with Flash, as an example of the flexibility, and we are finding that there are no limits." (Project manager, Web design and development agency)

Myth: Open source products contain all the basics. The products are evolving rapidly: Apache OFBiz, though currently geared toward enterprise resource planning (ERP), is ramping up the eCommerce capability, and Magento is little older than a year. Many commercial products have been maturing for years. Many of the core business tooling and site features — such as content management, search, and customer relationship management (CRM) — that midsize and enterprise online retailers and businesses expect may not be there (yet).

"I do not think [it] is the best eCommerce product, but it is a good development platform. In terms of elements that we were told were on the road map, our team went ahead and did them." (Vice president, eCommerce midsize multichannel retailer)
"The framework and the architecture are exactly what we need. There is flexibility to meet unique client requirements, while maintaining the core. But realistically the finished product is not quite to the scale of the midmarket. Some of it is there, but not at the level necessary. So we have had to build workarounds to meet the client needs." (Project manager, Web design and development agency.)

For our clients, I encourage you to read the full report here. In the report we have a set of criteria you should use to determine whether or not an open source solution is right for you — and, of course, also explore the other ‘promises’ as well in depth. And, as always, I also invite clients to contact us with further questions through inquiry.

While the open source eCommerce platforms may not be quite there yet, Forrester anticipates growing interest in eCommerce open source solutions in both North America and Europe based on discussions with retail and wholesale businesses exploring eCommerce platform solutions. This increasing trend is one that Forrester will continue to watch and evaluate.

We can only expect that the trend will intensify, based on further maturation in the products; evolution of enterprise services and support packages; as well as proven implementations at increased scale. Increased numbers of skilled and experienced resources and training programs will also help businesses take advantage of these open source applications. Commercial eCommerce applications, especially those focused on mid-market firms, will come under increasing pressure to provide value above and beyond that offered by open source solutions.

Thanks very much, Brian

twitter.com/bkwalker

Preparing to React to Rapidly Emerging Events: Tips for Your Online Business

by Brian K. Walker

If you are like me, you are following the developments with the H1N1 flu closely. Many businesses are seeing a spike in demand for products to help prevent spread of the virus, such as alcohol-based hand sanitizers and face masks - or to treat the pandemic should it be contracted, such as inventory of Tamiflu. Thankfully the severity of the flu appears to be far less than originally feared, though certainly that may change.

But these are certainly not the only large and emerging events that a business must react to. The Web channel offers a unique opportunity to react to emerging events both within the eCommerce channel and in reaching loyal customers. Hopefully there are few situations like the swine-flu – or other natural or man-made disasters – in the future. But by planning and preparing your business may be in a position to help those affected, drive a positive from a negative, and quite possibly experience some strong business as well.

Being prepared for the worst can make you even more responsive for great times, too -- for those profit-making opportunities that pop-up and which depend on quick follow-up and responsiveness.

So what can a business do to prepare for these events? Here are a few things to focus on:

• Resources. Have a team with the right skills and system permissions in place. For example this team may be comprised of a merchant, a copywriter, a marketer, a PR person, and a developer or two — all identified as a rapid-reaction team. Assignment to this team may need to be done on a rotating basis, especially during the holidays. Be sure to have critical resources at your technology vendors and marketing and PR agencies identified too, as well as know-how to reach them at their lake cabin.
• Process. Have a process in place — with clear decision-making and leadership — to react to events, approve plans, content, code, and so on. Have a means of communication with contact information sorted out up-front. Maybe this requires pagers, but more likely will comprise cell numbers, IM, texting, home numbers, and — today — Facebook and Twitter.
• Microsites. Be ready to launch and deploy a microsite quickly. A microsite will serve as a set of pages and products if microsites present a challenge with your current platform, prepare by developing a site with default content with a sample catalog and have your team prepared to modify it. Branding and overall user experiences should be very consistent.
• Landing pages. A cousin to the microsite, but possibly much easier and faster to deploy, a landing page with links to content, categories, and products may meet the needs of your audience. This may be a useful tactic in concert with the microsite to drive traffic between the sites as well.
• Content locations. Have content locations on the site ready to deploy. For example, a thin banner across the top of your pages (right under the global navigation) may be a great slot to use to direct site traffic. Do not overemphasize the homepage, as much of your traffic may never see it. Be sure you have content slots on category and product pages as well.
• Category management. Be sure your team knows the tools well enough to define new categories on the site and react quickly. Category naming may prove very important to customers finding the products.
• On-site search. Tagging, metatags, URL-redirects, binning, and content indexing will be critical to helping customers find the products quickly. A strong on-site search tool is key.
• SEO/SEM. Search engines are a key conduit for audience traffic. Being able to react to the terms and capitalize on the search traffic will be important.
• Marketing. Have a plan in place to review existing marketing plans and efforts, with the ability to react. This can save face, as a key promotion to an area hit by disaster or in the news could be embarrassing. (I will not provide a gruesome example, there are many.) At the same time, have the resources, skills, and authority in place to react to the event proactively through email, display, search marketing, and social media. A well-timed tweet on a timely and needed product may help.
• Establish key relationships upfront. Other key to planning ahead for situations like 9/11, the 2004 Indian Ocean tsunami, or Hurricane Katrina is by having relationships with qualified organizations, charities and service organizations established ahead of time. In some cases you may wish to prepare for a donation to be made through your site. Another means of support may be to link to other companies who have the products in-stock if you happen to run out, or offer free shipping for critical supplies. By having the processes we identified above in place, you may be better able to react to these opportunities in a quick and controlled manner.

I think it goes without saying that another key to reacting to these situations is to be charitable. This will pay long-term benefits in terms of consumer brand loyalty, employee morale, recruiting, and overall PR — and it also goes without saying that it’s the right thing to do.

Now, I do not want to be all doom-and-gloom, so I want to end on a positive note. These same tactics can be used to in context of a positive event too. For example, we have had clients who have had a product featured on Oprah, and the resulting traffic and sales opportunities are like an earthquake — but a positive one. Your team may only have hours or a day to react to the news that Oprah is featuring a product or destination, or the NBC Nightly News is featuring a story on a key product or category. These positive events may leverage much the same processes as above, and be good rehearsals as well for similar events down the road.

Bottom line, days and weeks are not sufficient when dealing with these events. By planning ahead, you and your team will know how to react swiftly and effectively.

Thanks, Brian

(If you have other ideas or examples I missed, please comment below.)

Forrester’s 2009 B2C eCommerce Platform Wave

By Brian K. Walker

Phew. After many months of analysis and hard work we are very pleased to have published the B2C eCommerce Platform Wave report. Clients can access the full report here: The Forrester Wave™: B2C eCommerce Platforms.

eCommerce Platforms Are In Demand

Our recent survey of our eBusiness panel shows 35% of companies are looking to replatform in the next two years. This is very healthy demand in an otherwise dour economic climate. Why? There are many factors, among them:

1. eCommerce is among few bright-spots for retailers and can be a way to offset diminished demand while closing retail stores, cutting retail staffing levels, and cutting back on inventories,
2. Wholesale-brands and CPG’s are looking to eCommerce as way to offset their wholesale-distribution channels’ lost sales and it is a much easier conversation around channel-conflict now when starting a direct relationship with consumers,
3. Established online retailers are dealing with aging site infrastructure and technology as they look to innovate around content management, product discovery (wide group of technologies here), and social commerce,
4. Integration to enterprise or legacy systems such as BI, CRM, ERP, OMS has become more and more important as the business matures and becomes a more significant component of the business, and contemporary eCommerce platforms are much more compatible with these requirements.

 

What You Must Do Before You Select A Vendor
Do not make the mistake of taking the top vendors in this, or any, evaluation and talking only with them. You must work to identify your business goals, brand and service differentiators, desired customer experience, realistic technology environment, operations capabilities and goals, and 5 year strategic goals before making a thoughtful and confident vendor selection decision. You will be living with this decision for 5 to 10 years. Forrester and others provide services in support of this effort. I encourage you to explore that.

 

Over 75 Vendors in the Platform Space

I am tracking 75 different companies in the eCommerce platform space. I know there are very likely to be others out there not yet on my radar. Many of them do not cater to the same clients – some focus on SMB, while others on B2B, etc. – but nearly all of them will say “yes” when asked if they can help an enterprise-class eCommerce retailer or brand with their eCommerce platform needs. We worked hard to focus on ten vendors who we felt focused on enterprise eCommerce businesses doing over $100 million in online revenues. We accomplished this through an RFI process and by relying on our own client-interaction data. Sadly, there were a few vendors who just missed the cut.

 

Key findings from the Wave Report

Art Technology Group (ATG) and IBM led the pack with their comprehensive eCommerce features, effective business tools, and flexibility. Fry and hybris represent strong solutions with varying models and key differentiators. Demandware, iCongo, Escalate Retail, and Intershop each represent unique solutions at different stages of maturity and are focused on midtier eCommerce retailers. MarketLive and Microsoft’s Commerce Server 2007 round out our evaluation.

 

Look beyond the surface

Now, as I say to our clients all the time, each of these solutions represents a fit for a segment of the global B2C enterprise eCommerce platform market. The wave reflects a specific filter we apply which we believe best reflects the requirements of an online retailer doing over $100 million in online sales, who has a typical enterprise integration scenario, and typical B2C eCommerce merchandising and marketing requirements. But, as you know, each company is different. (More on that below.). It is also important to look at the eco-system around a platform selection, such as system integrators, consultancies, and access to development talent when making these decisions.

Forrester Wave: B2C eCommerce Platform Solutions, Q1 2009

Some Thanks, A Real Team Effort

I want to make sure to thank a few people. Brendan McGowan for his attention to detail, focus, and for keeping me on track with the process. Nicole Lesperance for some great editing in the heat of the Q4 rush. Jens Kueter and Ad'm DiBiaso for the always high quality graphics, and Rebecca Anzalone and the production team for quick and accurate work on these large projects. Carrie Johnson for her seasoned, reasoned, and insightful editing and leadership during the process. And let’s not forget the vendors’ staffs who spent countless hours themselves responding to our requests for information, preparing demonstrations of their products against our specific and unique scenarios, traveling to meet with us, and in their (sometimes) reasoned and active discussions over why a score or comment was what it was.

 

Let's Keep The Conversation Going

As Jeremiah Owyang mentioned in his recent post about his wave on Community Platforms, our Waves are snapshots in time.  While extremely detailed and as comprehensive as possible, they cannot accommodate vendor changes which occur after the research is complete.  So, I'd welcome an ongoing dialog about this research, these firms, your own vendor selection, and other vendors that you have worked with. 

And to those vendors who were not included in this group of ten, please continue to engage. We often refer clients to solutions not covered in the Wave or our other research as appropriate and where we think there is a good fit.

For you process oriented folks out there, here is a link to some additional detail about Forrester's Wave product and our methodology.

A Reminder

And a reminder to clients, we love to help you with your vendor selection processes. We can provide a variety of input on the best-fit eCommerce technologies, including platforms. I look forward to talking with you about your needs. We work closely with many clients to help identify the right-fit solutions. The Wave tool also enables a do-it-yourself aspect by enabling you to create custom weightings. I hope that is helpful.

Best regards, and as I said, I look forward to the continued dialogue,

Brian

Note: Please leave comments, I look forward to reading and responding as appropriate.

 

Your Site On Drugs: The Hidden Cost Of Ad Blockers

by Brian K. Walker

Ad blockers have grown in use over the last few years as a plug-in for Web browsers. This is particularly easy for Firefox users who can easily add in the Mozilla recommended Adblock Plus add-on when downloading Firefox.

Usage stats are hard to come by. Unfortunately, Forrester has not been polling consumers with our own Consumer Technographics® product, which is otherwise very comprehensive. (I will work to change this.)

When I ask advertisers and retailers about this I often get a response along the lines of “Huh?”

So what is the big deal? A few ads get blocked, so what?

Well, based on a few informed conversations of late I think we are approaching the 5-8% range for consumers using ad-blockers. This starts to really count at that level.

There are three key reasons why this is important:

1. Missing impressions — the way the ad blocker works is that is does not render the ad in the browser. However, the ad was served by the ad server, content management system (CMS) or image directory. If you are selling or buying impressions this means you are either cheating or getting cheated. At 5% I think this starts to matter to the advertising buyer.
2. Many eCommerce sites use the same directories, or servers serving the ads, to serve merchandizing (or worse) instructional and service content. Often this is done to leverage      targeting capabilities available in the ad server, not available on the  legacy commerce platform.** This means sites are just plain broken on ad-blockers  — either the design looks terrible, or content the customer needs is just plain missing.
3. Many eCommerce sites are looking to monetize their traffic in other ways besides conversions. This  is especially true for high-traffic sites. This is bound to increase, and sites will be more and more impacted by a problematic ad-serving approach.*

Here is an example of Wal-Mart’s home page just by way of example:

So, what can you do about it?

1. Look at your site on ad blockers.How bad is it? Note what works and does not work, and programmatically address it with your team. ***
2. Short-term, change the directory path of served content that are not ads (and change the path of      ads at your own risk).
3. Longer-term, invest in stronger CMS and segregate ad serving from content serving and targeting. We have analysts who can help with this.

Let us know how we can help, and please comment on the blog with your thoughts on the issue.

Thanks, Brian

* For a good example of a site-design and execution which work well wither way, check the NYT site with and without an ad-blocker.

** Note that many of today’s leading eCommerce platforms support this capability. See our upcoming eCommerce B2C Wave report for more details.

*** While you are doing this also look at your site in multiple browsers (Recommended: IE 6, IE 7, Firefox, Safari, and iPhone-Safari) – oh, and order some things from your own site while you are at it.

If you are on Twitter, follow me at: http://twitter.com/bkwalker

MR. ROGERS’ GUIDE TO AN E-COMMERCE SLA

by Brian K. Walker

Over the years I have had the opportunity to be on both sides of the negotiations over service level agreements (SLAs) between eCommerce businesses and the hosting and service providers they contract with.  The two sides are obviously after different things, and have different goals. But as an eCommerce business what should you be looking for in an eCommerce SLA?

(Note: Want to jump to the punch line? Skip to the bottom of the post. Otherwise please enjoy our trip to the Land of Make Believe.)

First, A Trip To The Neighborhood

As a child a big part of my afternoons was Mr. Rogers, and I would get particularly excited when it was time to visit the Land of Make Believe. So, come with me won’t you? (Queue the trolley music and…)

Imagine many, many thousands of funny-looking hand puppets clicking away and shopping merrily on your site. Business is way up. You, as King Friday, are the benevolent ruler of this terrific business and Queen Sara Saturday (your CEO) is quite pleased. Things are going swimmingly and you and your merry customers are so pleased. You revise your forecast, call Henrietta Pussycat, your marketing director, and Daniel Striped Tiger, your merchandise planner, and say “Double down, we’re going places!”

Not A Happy Day In The Neighborhood

But just after you fired off an email to a million of your happiest customers and launch your big paid search buy and display advertising program things go terribly awry. First, Queen Sara Saturday calls and says her son Prince Tuesday can’t log-on to the site from his ski vacation in Vail. Then, just as you are checking into that, X the Owl, your customer service manager, hoots that his team is being flooded with customer complaints that the site is really slow and seems down. A quick check of Twitter shows that people all over the place are having big problems with the site. Finally, you jump over to your laptop. Sure enough, nada. Nothing. No site. 404. You immediately call Lady Elaine Fairchilde who is your key acct rep at the vendor you use.

Now What, Mr. Friday?

So, after much scrambling, and many conversations on the phone and IM with Ms. Fairchilde, your eCommerce platform service provider you are reassured everything is going to be resolved very quickly. You decide to wait it out, don’t panic, and reassure your management and team in turn. You even sleep soundly that night. But then when you wake up in the morning you find nothing has improved – the site is intermittently up and down, but when it is up it is slow. Something to do with a corrupt customer database and a problem with the SOAP service that manages the cart. But, Ms. Fairchilde explains, “No one told us about this campaign and this huge traffic spike. We were caught flat-footed.” This goes on for days. Your plan to double down has gone completely flat. Now you are sitting on a bunch of merchandise and have a bunch of angry, disappointed customers. So, you take the next most logical step and call your corporate attorney Peek-a-Boo Clock.  Your question, “How can we hold this company accountable? Can we recover some of the lost sales, or margin, we would have otherwise received had our site not gone kaput?”

Everyone Has Feelings

Here is where we need to get serious. “No more little ditties about how we are feeling” you say to yourself.  Mr. Peek-a-Boo calls to say, “I am afraid we are not in a very good position relative to the agreement and SLAs you reached with Ms. Fairchilde’s company”. They owe you nothing – except maybe a small percentage of the fees you own them for their services – in fact unless the site has been down more than 2% over the last 12 months, basically they owe you nothing.

The End Of Our Little Dream – What Can You Do?

OK – I apologize for the long setup. I could not help myself – I don’t have to go through editing with these posts (bound to change now). Now, let’s get to the punchline:

eCommerce businesses and the hosting and service providers they contract with are obviously after different things, and typically have different goals:

• The eCommerce business has a site and customer experience in desperate need of development or overhaul. The business leader has a goal and is accountable for getting the site up, operations humming, and transaction flowing - all tied to a revenue goal which was used to justify the project in the first place. And the next holiday-season is in the distance like a steam-whistle on a freight train running down-hill without any brakes. No time to waste.  No time to wrestle over the fine print, and the internal counsel is not often faced with questions over SLAs, so it really kind of gets skipped in the legal process.
• The eCommerce solution provider is trying to minimize risk and exposure. No service provider wants to provide an SLA that will risk their long-term recurring revenue.  Meanwhile, they also want to get the deal done. The sales people want to book the business, and the professional services people hear the same whistle coming. Meanwhile the legal counsel is much attuned to questions over SLAs and is motivated not to commit to anything.  They have all the arguments prepared.

All this often leads to SLAs being pretty “boilerplate” and based on the vendor's paper, with little real security for the eCommerce business. Hopefully all goes well, and this never comes up.

But of course the point of a good SLA is to create accountability and outline a remediation should anything go wrong. While you need to consult with your legal counsel on specifics, here is a breakdown on some common things to look for in SLAs from an eCommerce business and operations perspective:

• Ensure you are compensated for poor site performance. Remediation should include a sliding scale of performance problems, including downtime as well.
• Be aggressive on the thresholds you are willing to accept. 98% availability means roughly 14 hours down a month of unplanned down-time. I don’t think any of us would accept that. Go for 99.99% but we willing to back-down to 99.5% – that’s ~3.5 hrs of unplanned down-time per month
•  Remediation should include some percent of lost sales or gross margin. Compensation based on some sliding scale of the fees you are paying is not putting teeth into it. Develop a scale whereby the vendor has to compensate for lost sales or gross margin if things get bad enough.
• Insist on third-party monitoring tools to ensure transparency. Services from Keynote Systems, Gomez, or Tealeaf may be appropriate, and there are many lower cost services out there who can provide automated monitoring. Tip: Look for automated email or SMS alerts to notify your team of problems. Do not rely on vendor-supplied reporting, no matter how much you trust them. Their reporting may not be that good.
• Consider an index of sites to use as a benchmark. This is a way to not accept mediocrity, but also help the vendor mitigate problems outside their control. Using the same monitoring services used on your site, ping key benchmark sites and create an index. Alternatively you can use an index such as that provided by Keynote.
• Do not allow the vendor to maneuver out of responsibility for your sites performance and stability. If you ask for something, even a customization, it is their responsibility to QA and roll-it-out to production in a responsible way. They need to take ownership. If you add a third-party piece of JavaScript, same thing. If you send them a flash file or some HTML, same thing. Watch out for these exceptions in the SLA.
• Close force-majeure loop holes, and insist on failover and redundancy. Stuff happens, and sometimes it is really bad stuff which I prefer not to belabor, but in today’s hosting and cloud computing environment there is no excuse for a vendor serving the mid to enterprise eCommerce solutions marketplace (or high-scale SMB for that matter) not to be prepared for the worst. Insist that failover and redundancy guarantees be in place, and that the processes, procedures, and architectures are documented, expressed and explained.
• Be proactive with your vendor on your campaigns and traffic forecasts. Don’t make the mistake King Friday made: communicate with the vendor on a regular basis. No one wants a problem and the vendor can only be prepared if they are informed.

I hope you enjoyed this little tale and that some of these pointers are helpful, and as always, let us know how we can help.

Please add to the comments here on the blog to share your perspective on SLAs from the customer and vendor perspectives.

Thanks, Brian

On Twitter? Follow me at:  http://twitter.com/bkwalker

Or please feel free RT:  @bkwalker Guide to eCommerce SLAs:  http://budurl.com/j342