We're a strategic management and technology services consultancy.
We advise our clients on critical technology investment decisions. Our clients work with us whether they're considering a private equity investment, developing a global trading platform or want ideas on how to make their products more competitive.
Electronic and Algorithmic Trading, Risk Management and GreenIT. Read our articles and case studies.
Below I’ve listed 10 “facts”, “myths” and “suggestions” that you should be aware of to improve your odds of project success.
1. Fact: No one can reliably predict the future. The most obvious marker of project failure is overrun time and costs, which to an extent, require us to predict the future. In essence, we begin projects with little hope of predictable success[but this doesn’t have to be the case]. This leads to the next 2 “myths”.
2. Myth: Experts are good at setting estimates. Experts are great at determining the details involved in getting something done, and qualitatively describing the complexity of a project. However, they are no better than non-experts at predicting the end of a project. So, leverage experts to define details and leverage “everyone” and data-driven methods to determine the ETA(estimated time of arrival) and confidence intervals.
3. Myth: Professional Project Managers are good at setting estimates. Like the experts, they’re no better than non-experts. However they have tools for collecting estimates, challenging estimates and following up… so Leverage their tools and techniques.
4. Fact: Project Complexity grows exponentially with degrees of freedom including: people involved, number of tasks, system size, time etc. Simply adding individual task ETAs (Estimated Time of Arrival) will not give you a reliable picture of the overall ETA.
5. Myth: “Under-committing and over-delivering” is a smart project management tactic. Wrong! I think people are too smart to fall for this hack. I’ve found that the best people typically *always over-commit* and *sometimes over-deliver*. I’m not aware of any great projects that have begun by setting unusually low expectations, followed by spectacular delivery.
6. Myth: Good Project Management Implementation leads to successful Projects, ie On-time, On-budget. Wrong! I’ve heard Project Management credited with both success and failure of projects. In my experience, “people” are the main reason why projects both succeed and fail. check out http://blogs.zdnet.com/projectfailures/
Now for a few things you should do…
7. Suggestion: Whatever you do —- always collect and save data about estimates! The data will allow you to go back,analyze and test your more quantitative approaches. This brings me to the next suggestion…
8. Suggestion: Project Management and in particular Estimation must be quantitative and data-driven. There are many techniques including monte-carlo simulations, options pricing, portfolio theory, prediction markets etc. Determining the right tool for a project might take some trial-and-error but well worth it. I’ve often employed simple models where I break down the data along various dimensions like person, duration, complexity… You might simply begin by playing around with the data in a spreadsheet.
9. Suggestion: Keep it simple. Do use standard tools and best practices but don’t let that become a project. Too many reports, people, charts, tools etc is no good and only adds to complexity.
10. Suggestion: Make stakeholders/clients part of the process. They should share accountability for the plans and therefore share your pain when targets are missed. Typically, they not part of the process or largely focused on providing requirements. Take their input and leverage their experience… they’ve done projects too:)
One more: Stock your PMO or Project Board with people who will actually give you meaningful feedback… Not just Project Managers, Stakeholders or people you want to score political points with:)
I’m currently researching a few products and techniques and promise to post my findings.(photo via darkmatter)
Almost more than anything else, I’ve always enjoyed identifying or
creating something new and useful. I’m also a firm believer that the
opportunities to do so are often right in front of you. Many folks jump on the opportunity to innovate when starting a project from scratch, but readily miss on-going opportunities.Consider client feature requests for your system. Typically these are put on a long list of possible enhancements, later prioritized, then perhaps implemented and released. While I applaud the focus on your customer, I can’t help but think that perhaps you missed an opportunity to discover something innovative. Given this situation, a basic approach for discovering a better solution requires that you ask why the feature is needed and what they’ll do with it. Don’t fall into the trap of giving up the questioning too soon.
To illustrate my point, I’ll walk you through an example. A client of your trading system requests that you provide the ability to sort the PNL column by the absolute value in descending order. A really clever developer takes this requirement and overloads the grids sorting function so that any column can be sorted this way. Job done? Technically, yes. Brilliant? maybe. Innovative? Nope!
To my point, you should follow up with some questions. Below is a simplified version of a hypothetical conversation you might have with your client.you: how would you use it?
client: first thing in the morning after my coffee, I fire up the trading system and sort by PNL.
you: do you change it?
client: rarely. it’s the most important value for me.You could stop here, and simply sort that column by default, but there could be more.
you: what are you looking for in the PNL column?
client: basically, any PNL exceeding 20% of the total value of my portfolio.
You could stop here, and highlight those rows that match the criteria… but why stop thereyou: is it always 20%?
client: sometimes I set it lower or higher depending on what the risk management guys tell me.
you: what do you do with the rest of the portfolio? those positions that do not exceed 20%.
client: nothing really. there are too many of them. in fact the ones scrolled off the screen never get my attention.
Now we’re getting somewhere… maybe you remove the clutter and only
display those rows. BTW… I rarely see this done in trading systems.you: Now that you’ve identified the 20% movers, what do you do next?
client: I look for another security that’s highly correlated but hasn’t experienced the same movement.
you: How do you do that?
client: I use this nifty Bloomberg screen, type the sector and some other criteria and bam I have a few stocks.This is a possible Integration point. You’re client might not ask for this, but you might suggest a way to automate this bit by integrating the two systems.
you: do you then buy or short the stock?
client: exactly. I expect it to move so I try to pounce on the opportunity.
Perhaps you can automate the trading or alert the client of an opportunity and have her simply click once to trade.you: cool. This is great information. I’m going to get back you tomorrow with a few possible solutions. Then we’ll talk about prioritizing it and when it could get done. What do you think?
client: sure… but I’m dying here without the ability to sort that column.
next day…you: here are the ideas I promised… I also have a general idea of how difficult they are to implement. It could help us determine which solution to take.
You propose a few solutions ranging from simply hacking in the sorting functionality to the more innovative solution of automating this process. Automating the process might change the game… Innovation!client: let’s work in the sort function first so I can start making some money. However, automating the process seems like the way to go. Damn you’re brilliant. Flush out a proposal that we can show to management.To sum it up, here’s what you should do.
- Assume that the feature or isolated need is part of some bigger need.
- Ask why they need the feature or enhancement
- Ask how they plan to use it
- Drill into their process to discover the underlying motivation
- Each step of the way, think of ways to solve the problem
- Ask yourself if you can do more… you know you can.
- Finally, share your ideas with your team, colleagues and most importantly, your client
Now go back to that list of small client requests and start a dialogue with them.
Good Luck(photo via iLoveButter)
from Wall Street and Technology
I generally disagree with this approach. It’s precisely during times like this that you find opportunities to pull ahead. While everyone is simply focusing on efficiency and compliance, you should take advantage and make some strategic investments. Get them while they sleep… I’m looking forward to the followup articles by WS&T.
Update: Jan 9,2009 - WS&T has a set of video interviews discussing Capital Markets outlook for 2009. Focus on efficiency, “running the business”, rethinking market microstructure and behavior, and more. [video]
Update: Dec 18, 2008 - Read this interesting article about what Google has been doing in spite of the recession including new ad approaches for GMaps, Youtube, iPhone, News and Finance.