Handling 3rd party vendors on SCRUM projects

A friend of mine recently asked me what the best way was to manage a 3rd party vendor on a SCRUM project. (SCRUM is an Agile Methodology for those of you who don't know)

Here's the advice I gave him:

Get the vendor relationship off to a healthy start:

• Make sure there's an open, honest and safe channel for the vendor to communicate with you.
• Be as specific as possible with what you require from them. Try and be results focused and not activity focused. Here's a great article by Marc McDonald on the merits of results versus activity management.
• Urge vendors to give commitments that you can rely on and not commitments that you may want to hear.
• As a rule of thumb question the assumptions behind an estimate and not the estimate itself. This way your vendor is encouraged to use their expertise while protecting their self esteem (improves their perception of safety).
• Try and use the language of 'confidence' when considering estimates. For example, "The vendor is very confident in a 5 day delivery" or "The vendor is not very confident in a 5 day delivery due to an issue with XYZ". This way you're less likely to set unrealistic expectations with stakeholders.
• Ask your vendors to escalate potential issues early and often... better to handle things early & cheap than late & expensive. (See my previous blog article on this)
• Work hard to empower your vendor... especially since their performance has direct impact on your reputation. In other words, your reputation will very likely take a knock if they flounder.
• Prioritise the deliverables you're expecting as much as possible. Be clear about Must Haves.
• Compromise on Scope not Quality where possible. Don't make them deliver things if it can't be done properly. For example if you're time pressured, negotiate with scope.

Integrating vendor deliverables with your SCRUM project:

• Create a user story card for their deliverable and place it in the product backlog.
• When the iteration arrives be sure to schedule integration and testing tasks in the sprint backlog to ensure the vendor deliverables are truly DONE by the end of the iteration.
• For planning poker, Limit the task estimates to just those your team will need to carry out. I.e. exclude any work your vendor will do as this will distort your view of velocity (... thanks Vasco)
• Invite your vendor along to attend the sprint planning meeting for the relevant iteration. A public commitment to the team to meet a deadline is always more powerful than a contract ;-)
• Be sure to regularly ask vendors how they are tracking and to confirm their delivery estimates (weekly should be ok).
• If there are dependancies on the vendor's work be sure to assess the impact of delays. Communicate the potential impacts to stakeholders early.
• The vendor should also attend the sprint review to see how their deliverable has been integrated and to clarify anything should the need arise.
• It's also ideal to have your vendors attend the relevant sprint retrospective to get their input.

Vendor Selection Tips

What's a reasonable vendor selection process?

Selecting the right vendor starts with knowing what you want and then finding the supplier that will get you what you want. Here's a breakdown of what this might entail:

Step 1) Understanding the requirement
Step 2) Generating the options
Step 3) Analyzing the options
Step 4) Selecting the best option
Step 5) Getting the decision to proceed

Informal vendor selection (Fast track)

Use this approach if you need to turn around a decision really quickly and you have the remit or necessary sponsor support to keep it informal.

Step 1) Understanding the requirement

It's pretty important that you have an idea of the desired outcome before even considering vendors. A standard gap analysis is a great approach to getting to understanding the problem. Work out where you want to be (i.e. the end state) make sure it's ecologically sound. That is to say that the outcome is beneficial in some way to all concerned! Invest the time to articulate exactly what you want in the form of a problem statement(perhaps reinforced by a statement of what you don't want). The principle being that your chances of getting what you want dramatically increases if you know what you want ;-). It helps to use the problem statement to identify your selection criteria (see Step 3 below)

Try to have the target budget identified. This may need to be challenged later depending on the criticality of the desired outcome.

Step 2) Generating the options

Use your problem statement (created in Step 1) to identify possible vendors. You may want to contact your procurement department to check whether they have any preferred suppliers to hand. (Note: You may need to be sensitive to existing supplier relationships). The next stop should be your immediate professional network followed eventually by speculative internet search. Generate the list of potentials bearing in mind that time is of the essence.

Contact them by phone at first to weed out the obvious mismatches. Once you've got a short list together send them your problem statement via email (or mail). Follow this up with a phone call explaining your time constraints and openly request their support. Be sure to share any information fairly across your vendors. (e.g. an answer to a question may need to be shared with the other vendors)

Set-up a demo day for all your vendors.

Step 3) Analysing the options

Invest in getting a few key assessors together that will balance your judgement. The pay-off will be that your selection is less likely to be overturned later on. These additional perspective will ensure that your blind spots are covered.

Create a score card with the selection criteria down one side in rows and the vendor options in the columns for assessment. The intention is to score each criteria out of five for each vendor (where 5 = 'excellent fit for purpose' and 1 = 'unfit for purpose').

Apart from the usual logistics (meeting rooms, diaries etc.) make sure you have briefed your assessors on the problem and the instructions for filling in the score cards.

Step 4) Selecting an option

Make sure the vendor demonstrations focus on addressing the problem statement. Collate the score card results. Add the scores up for each vendor and calculate the percentage of the maximum available. You can express this percentage as 'the estimated percentage match to requirement'.

Offset these percentages against the cost profile for each vendor. Try to identify the vendor with both the highest percentage match and the lowest price.

Discuss your findings with the other assessors (make sure you have consensus).

Step 5) Getting a decision to proceed

Present your findings to the decision maker(s) along with your recommendation. Make sure the next steps are agreed (i.e. what you need from them to proceed) this could be an authorising email or Purchase Order. This will depend on the Inform your vendors of the outcome and try and provide as much feedback as possible.

In part two we'll look at fast tracking this with little more formality.