LDSAudio.com VP Marketing Jim Ericson has been making in-house versus outsourcing decisions for every aspect of marketing tech from email services to Web analytics for almost a decade. Back when he was one of the very first marketers for now-megalithic MyFamily.com Inc, "I was pitched on a number of services that no longer exist. And there are still bottom feeder marketing services out there making money while not providing much value."
So, how can you make the decision to bring marketing technology in-house or outsource to a vendor? Ericson boiled down the decision to three basic questions:Question #1. Does it routinely touch your customers?
As marketing technology evolves, many services become commoditized. But if you're paying commodity prices, then your vendor isn't investing in the world's greatest customer service -- for you or for your customers.
If the service touches your customers on a regular basis, and you can't be certain the vendors in the field will take your customer relationships as seriously as you do, then don't outsource unless it's the only way to get the job done on budget. Example - telemarketing is rarely as effective when outsourced, but you may have to outsource for off-hours and at-peak overages just to get the calls answered.
If the service involves nothing but technology, it's easier to outsource. For example, Ericson says, when Ancestry.com started its affiliate program, "we ran it for three months, and then our accountant came to me very upset and said, 'Do you really expect me to cut 3,000 checks this month?' So we outsourced the architecture but maintained the relationships. If these people are really important to your success, it's important that *you* manage the relationship."Question #2. How core to your business is the service?
If both your brand and your bottom line are at stake, then you need to both hire the absolute best vendor possible and budget accordingly or you need to bring the service in-house and hire accordingly.
Sometimes it can be tough to get the right budget for outsourcing though, due to the perception in many C-level executives' minds that Web-related services should be extremely low-cost or even free.
Example: A few years ago email was promoted so heavily as a low-cost to no-cost alternative to snail mail that getting the budget to hire a best-of-breed vendor may be hard. The same holds true for search engine optimization services, for years known as something you could "do yourself" or that gets tossed in as part of a standard Web site design job.
The key is to truly estimate every cost you'll have if you decide to bring the service in-house at the level of competency your business requires. How many staffers will you need to run the new service? What type of equipment or software? How about training? And what happens if you take your eye off the ball and the service is badly run? What's the cost of screwing it up in-house?
Example, some magazine publishers have considered owning their own printing plants. But, content and ad sales are core to their business -- not paper and printing. So they leave the printing to the printing companies where print management is the core of their business. Question #3. How much work does the service really entail?
If you're considering a build-it or buy-it decision, create an RFP that's sent to both groups of people (yes, including your in-house programmers). Include the basic specs and dated expectations for launch, as well as maintenance schedules, upgrades, and customer service requirements.
If your in-house programmers are getting one of the RFPs, be sure to ask who will be the project manager and what other projects they would have to delay in order to take on the extra workload.
Double all programming estimates from both sides in terms of time and cost, plus headaches. Vendor-promised upgrades rarely materialize completely on time, on spec, and bug-free. And just like home improvement projects, in-house programming is always harder than you thought it would be.
Worth noting: If the service exists in the outside world and it's been around for a while, question your eager in-house programmers' motives. Do they want to reprogram the wheel because they think it would be a neat-o project or is the existing vendor technology so far off your RFP's requirements that lots of programming is needed? Far too often the answer is the former.
On the other hand, if your IT department (or other involved departments) beg you to outsource because they are already too darn busy, be forewarned that outsourcing doesn't equal no work. You'll need IT's help to tie back-end systems together, replicate databases with secure data, and even hunt down answers when there's a problem and all your vendors are finger-pointing at each other. If your IT team truly is too busy to help with project management, you'll need to budget to hire a consultant or even possibly a full-time staffer to pitch in.
Examples: In-house IT and Web departments are infamous for dropping the ball when it comes to loading up newly search-engine-optimized pages. Marketers spend thousands on pages and site structure formats that never get used.
If you've come to your decision and the answer is "outsource it," then be sure to watch out for Ericson's Seven danger signs of a dud vendor.Danger Sign #1. Are they scalable?
What if your marketing test runs out of control? For example, perhaps you test PPC advertising and it works so well that one month you're doing 500 keywords and the next month it's 50,000? (Yes, it happens.)
You could find yourself under contract with a vendor who no longer can handle the size of your account. It's happened to Ericson with email services when one company's permission list ballooned unexpectedly quickly from a few thousand to over a million.Danger Sign #2. Is their terminology misleading?
Ericson notes many email service providers' sales reps use misleading terminology when they pitch the service. For example, anyone who promises 99% email delivery rates probably is being misleading, because spam filters stop 20%-40% of all permission mail these days. (Chances are they use the term "delivery" to indicate messages sent minus bounces, which isn't actually end-delivery.)
Another typical misleading promise: Some search optimization firms promise to guarantee you rankings for a certain number of terms. The offer may be worded in such a way that potential clients think they can get high listings for any term they desire.
However, guarantees are impossible in optimization. (No search engine is under any obligation to list a site in their organic listings.) About the only things you can be pretty sure basic SEO will help you get listed under are your company/brand name and your Web site name. That's a relatively unimpressive guarantee because it truly is something you can do yourself (unlike being listed for more competitive terms).
So ask all vendors to include a glossary of terms along with promises so you're sure they mean what you think they mean. Danger sign #3. What's the guarantee actually guaranteeing?
Is the guarantee just marketing window dressing to close the sale, or does it really minimize your risk? Be sure to get a definition of precisely what's being guaranteed. For example, ASPs are infamous for guaranteeing 99% uptime, which sounds impressive until you realize it allows them to be down for more than seven hours per month.
Next, find out what you get if the promised deliverable is broken. Will you get that month's service free? Will the entire contract be null and void and you're free to seek a new vendor? Will you be reimbursed for any extra costs or losses you incurred as a result of the outage or problem?
Generally marketers blindly hope the latter is true, even though it nearly never is.Danger Sign #4. Who owns what?
If the vendor is touching your customer data in any way, for example credit card processing, powering a shopping cart, generating leads, etc., ask "who owns the customer?" Do you or do they? Who owns the records, plus who has the right to contact that customer in future?
If the vendor is powering any part of your site, newsletter, advertising, or database, be sure to ask who owns the content. Do you own the copyright on articles a third party has written for your email newsletter? Do you own the copyright on the new site pages an optimization firm may be hosting for you? What about graphics, photos, voiceovers, or other creative material in ads?
Don't forget to also ask, “What are the provisions for backup?" (Especially when working with ASPs -- if they crash or go out of business, does all of your data get lost?)Danger Sign #5. "We're growing!" Bragging
Growth may stink. If your vendor suddenly takes on a heaping of new clients, what happens to your account? Will your account reps be stressed because they have too much work? Will you suddenly be relegated to less-experienced staffers because you're not the big fish anymore? Is there a danger the CEO's head is focused on new client pitches versus perfecting service to maintain existing client relationships?
M&As and venture-backed funding are also warning signs. If a vendor brags they just got a million dollars in financing, you should wonder if they'll be so busy reporting to their new VC board members that they forget your priorities. Or they'll just be too busy dealing with the politics and business strains M&As always entail for at least the first six months.Danger Sign #6. No references
If a sales rep can't give you any references for any existing clients because he or she claims "We're under NDA" or "We're still in beta," run in the other direction. Every reputable vendor must have some clients you can talk to, and some give discounts (especially to beta clients) as a way of ensuring this.
If you yourself agree to become a beta client, remember it's a lot more work than a normal vendor relationship because things are guaranteed to be buggy. If you're outsourcing to save on staffing or time, becoming a beta client is the exact wrong way to go about it.
Quick tip: Always ask for a former client reference as well. All vendors who've been around for more than a year have former clients. There's no shame in this -- a former client may have changed their business model, executives, budgets, or plans. The fact is, former clients exist and you have a perfect right to demand to interview them as part of the RFP process. If a vendor says no, then that's another danger sign.Danger Sign #7. Outsourcing
Due to the revived economy, even the vendors are outsourcing. The programming or services you thought were being provided from an office in Akron may in fact be handled by moonlighters, work-from-home freelancers, and /or workers in another country.
If you have privacy issues -- for example, the vendor is building something that will give you an invaluable specific edge over the competition, or the vendor is housing data that's critical to either you or your customers -- then you can't use a vendor who is outsourcing anything.
Example: Many companies have been surprised to see their Web redesigns show up as part of an unknown subcontractor's portfolio on elance. Don't let it happen to you. (Or if it does, negotiate a discount up front.)Useful links related to this article
Cartoons of program development from the programmer's point of view (free and hysterical if you've ever done a software development project): http://www.hjsys.com/hjsasp/gn02.cfm?SI=5481565444152&ID=18
LDS Audio - a Division of Deseret Digital Media http://www.ldsaudio.com