If you've been paying attention, you know that I rail on about the problems of fixed-price CRM projects while endlessly promoting Agile engagements for CRM systems.Why? When you're trying to optimize a CRM system, you're not getting a pre-existing commodity. You're developing a customized system that helps you achieve your specific business goals. It should leverage (or provide the foundation for) organizational change. In other words, you're not buying a product: you're buying into a process.I'll even argue that if all do is buy "an implementation," then you're missing half the point of CRM. If all you get is "automation," then you'll just be doing the same old thing at lower cost, instead of growing the top line.But enough with the polemics.When the Boss Says, 'Forget Agile; It's Fixed Price!'For many, the biggest issue with agile projects is not knowing what you're going to get for your money. If you don't trust the project team and the vendor, agile can indeed look like a license to steal.The best answer to your boss' concerns: Don't even think about starting a project until you find a project team and a vendor that you dotrust. But that's probably just crazy talk. In my experience, fixed-price CRM projects are more likely to be complete in name only, with unhappy users, than they are to be satisfactory to users, with unhappy budget-holders.Analysis: How Much Should You Spend on CRM Software?(Full disclosure: I ran fixed-price Department of Defense projects for years, with only one small overrun. Despite the popular perception, Air Force procurement processes are pretty tight.) In CRM projects, clients tend to demand fixed-price projects without actually preparing for them. In response, vendors often use three strategies for "fixed pricing"\u2014providing only cookie-cutter CRM implementations, selling just a bucket of hours (that is, not signing up to completing the deliverables), or starting engineering change order (ECO) negotiations almost from the outset. To prevent this from happening, there are a few things to look for in the vendors, as well as a few behaviors you need to follow along the way. Here's a quick rundown of how things ought to work.Step 1: Before Any CRM Project, Do Your HomeworkThe RFI, RFP and RFQ drill works well if your team really knows the details of the business and technical requirements. Unfortunately, spec-writing teams tend to make three classic errors:They simultaneously under-specify (with too many silent assumptions and incomplete information) and over-specify (with too many product-specific stipulations).\nThey're vague (or even silent) about acceptance criteria for features. A spec without a test or a clear go\/no-go threshold is a wish-list, not a spec.\nThey oscillate between the dweeby features, the business rules, and the overall goals.Commentary: Careful: Your CRM Integration Consultant May Not Know AnythingEverything needs to be translated into pure system attributes and features. It's pretty rare for internal staff to really succeed with writing the specs on their own. Instead, bring in a consultant who will be excluded from the bidding to help you set CRM project priorities and write an iron-clad spec of what you'll need\u2014and only what you really need.Step 2: When Short-Listing, Pit CRM Vendors Against Each OtherAs vendors respond to your RFI, RFP or RFQ, look for more than just good answers. Does the CRM vendor proposal include rules of engagement for the project? If there's no guidance and dialog during the bidding process about the how the project will be managed, then friction is likely to develop later on. There are just too many places for mismatched assumptions.At some point, you'll eliminate some vendors. Identify which of the disqualified ones you communicate best with, and then engage them in a short but critical task. Make it clear that they aren't in the running, pay them for their time and ask them to do two final tasks:Apply Value Engineering to the other vendors' proposed approaches. By ruthlessly applying the 80\/20 rule, there is always a way to satisfy the core business requirement with a smaller level of effort.\nSharp-shoot the contract, looking for areas where you could get severely hit by ECOs and loopholes.During your short-listing cycle, modify your specs and contract to avoid the problems that the out-of-the-running vendor has spotted. Even if you pay the vendor $1,000 an hour for this task, you'll still save a bundle overall.Step 3: Pay Attention During the CRM ImplementationIf the winning vendor is smart, it will precisely specify what it will (and won't) deliver in the Statement of Work (SOW). Any change to the SOW will probably require an ECO\u2014each with its own price tag. This will be true even if the following happens:You didn't understand the consequences of the specified deliverable. For example, the SOW might say "20 triggers and 17 workflows," but you may have no idea whether that functionality will be satisfactory to your users.\nYou discover bad assumptions, additional requirements, or plot complications\u2014or you simply change your mind\u2014at any time after the contract is signed.\nAll the itemized deliverables do not solve your system problem, make your users happy or achieve your business objective.Also note that the SOW may contain a line item for an explicit number of hours of project meetings, as well as another item for project management. If meetings or your decisions\/approvals take longer than the allotted time, the vendor is within its right to issue a change order.\nCommentary: Why CRM Implementation Is So Political\nMore: 5 Ways to Send a Custom Software Project Off the RailsDuring serious fixed-price projects, meetings and the sequence of tasks will not be under your control. At times, it may feel like you're working for the vendor: You'll receive action items, decision deadlines and resource requests. In particular, you'll need to complete your testing before features can be fully deployed. The vendor has to be running the show if it's going to deliver on time and on budget with quality.In other words, be prepared for the vendor to say "No" fairly often. If this all sounds a bit adversarial, it is. The vendor's fixed price commitment is to delivering only the items you put in the spec, not to making you happy.This won't happen, though, if you prepare your team to the real requirements of fixed-price engagements. It may not be easy, but the short-term pain will result in long-term gains that make users and executives happy.David Taber is the author of the new Prentice Hall book, "Salesforce.com Secrets of Success" and is the CEO of SalesLogistix, a certified Salesforce.com consultancy focused on business process improvement through use of CRM systems. SalesLogistix clients are in North America, Europe, Israel and India. Taber has more than 25 years of experience in high tech, including 10 years at the VP level or above.Follow everything from CIO.com on Twitter @CIOonline, on Facebook, and on Google +.