Decision Tools: Strategy Sieve
How to Choose Between Options
This is the fifth in a series of articles on tools helpful to leaders in making hard decisions. This week we are looking at the Strategy Sieve.
What It Is
According to Wikipedia, a sieve is “a device for separating wanted elements from unwanted material” — and that’s what the Strategy Sieve does. You “pour in” a bunch of potentially good options, shake things up, apply some pressure, separate out the wanted from the unwanted, and determine the best decision for moving forward.
It provides a structured approach for comparing alternatives against criteria that are most relevant to the decision being made. The Sieve is usually constructed of a series of “filters” or collections of related criteria so that you can easily see how different options compare across meaningful dimensions.
Mechanically, the Strategy Sieve is a workbook made up of spreadsheets for each filter and a summary of the results across filters.
The Sieve can be used by an individual in making an important decision, but it’s real value shines when used in a team process, typically involving 3–10 people who have an important stake in the decision being made. When used in a group setting, the result is higher levels of confidence, clarity, and consensus in the decision being made.
When To Use It
The Strategy Sieve is most helpful when facing a decision that must balance multiple unrelated types of criteria and where the group involved in the decision can’t easily reach consensus on the best path forward.
How to Use It
There are two aspects of “using” the Strategy Sieve worth considering:
- How to build the Sieve
- How to use the Sieve to make a decision
Let’s look at each:
Developing the Strategy Sieve
I think it’s helpful to break down the Sieve development process into 5 steps:
- Identify the decision to be made.
- Identify and clearly distinguish known options.
- Identify “filters” — key themes to shape the decision.
- Populate specific decision criteria within each filter.
- Build out the workbook.
Let’s look at each of these in turn.
- Identify the Decision
You wouldn’t be at this stage if you didn’t have a decision to make, but how clearly have you spelled it out? Specifically you should identify:
- What problem are you trying to solve or what opportunity are you trying to capture?
- How will you know when you’ve solved the problem or captured the opportunity? What metrics or indicators will you use?
- What has caused the problem or created the opportunity? Has something changed?
- How are you currently trying to solve the problem or pursue the opportunity? What indicates that it isn’t working?
- What constraints do you have in solving the problem or pursuing the opportunity?
- What capabilities and resources can be applied to solving the problem or pursuing the opportunity?
- Who is negatively impacted by the problem? or Who will benefit the most from the opportunity?
- Is anyone relatively benefitting from the problem? or Who is likely to suffer the most from failure to capture the opportunity?
- Who is likely to be involved in solving the problem or capturing the opportunity?
- Who is the final decision maker?
- Who else is likely to influence the decision?
2. Identify Options
Specifying the decision in the detail outlined above will probably help you identify potential “right answers” for the decision. I find it very helpful to spell out the details on each option to clarify what is different between the different options. This can help you identify additional options, but will also help you as you walk through the process of using the Strategy Sieve to make the decision.
One way that I like to do this is to create a sort-of “future baseball card” for the option. If you were to summarize the key facts and highlights about each option, what would you say? It’s most helpful if you use a consistent set of questions across all the options and put them in a matrix.
As you do this, you also will likely find that some options you identified are actually very similar. It’s a good idea to try to get to a set of options listed that are mutually exclusive and collectively exhaustive (MECE), but often that’s not practical — there are often options you want to consider that overlap with each other. That’s okay. You will want to end up testing two to seven options, but some may get eliminated in the first step of using the Sieve, so having up to ten options may be okay at this point.
Here’s an example:
3. Identify Filters
Once you’ve identified options, the real question is “how will you decide between them?” Another way of asking this question is “what would have to be true for an option to be the right choice?”
Based on the work you did in defining the decision, you know the indicators of a successful decision, the constraints you’re working within, the resources you have available, and who cares about the decision (and what they care about). Based on this information, you should be able to identify three to five broad criteria that must be met, for example:
- It must solve the problem or capture the opportunity the decision is all about (e.g. “Revenue Growth”)
- It must be workable within the constraints and with the capabilities and resources available (e.g. “Leverage Existing Capabilities” and/or “People/Budget”)
- It must satisfy the end beneficiaries (e.g. “Student Needs”)
- It must deliver value that matters to decision makers and influencers (e.g. “Company Value”)
Each of those broad categories will become a filter in your Strategy Sieve.
4. Populate Decision Criteria
Under each filter you want to identify 5–15 specific criteria against which you can evaluate and compare the different options. Where possible, these should be quantitative and measurable, but more qualitative criteria are also valuable.
For example, for Company Value, you might identify six specific criteria:
- Revenue Growth Rate
- Total Market Size
- Operating Margins
- Investment Required
- Positive PR Value
- Brand Alignment
5. Build Workbook
The prior three steps give you all the information you need to build out a workbook made up of one spreadsheet summarizing the options, one spreadsheet for each filter, and one spreadsheet (linked to the filter sheets) summarizing the results. In the “Related Resources” section at the end of this article I include a link to a Google Sheets template that you can copy for your own use.
The spreadsheet summarizing the options would look very similar to the matrix of options shown above under step 2.
Each filter spreadsheet would have a column for each option and a row for each specific criteria:
The summary page would bring all the filters together into a single view:
Using the Sieve to Make a Decision
Once you’ve developed the Sieve, you can put it to work in making a decision. This works best in a group setting, so that’s what I’ll describe below.
There are five steps involved in using the Sieve in a group setting:
- Refine the options.
- Agree on the filters and criteria.
- Walk through each filter.
- Consider the results.
- Make the decision.
It is best to distribute the detailed definition of the decision (from step 1 above) and the workbook containing the options and filters (from step 5 above), along with an explanation of the process being used (for example, a link to this article) to everyone participating a few days before the sifting session.
The entire process can easily take a full 8 hour day. It may be best to split this across two half days, perhaps with a group dinner in between to decompress and reflect.
1. Refine Options
Everyone should have reviewed the decision being made and the options being considered before the session begins. Ask them to come to the session with questions and concerns. Set aside the first 30–60 minutes of the session to discuss, adjust existing options, and potentially add some that had previously been overlooked.
For the rest of the process, you want to limit the number of options being considered, optimally to 4 to 6 different options.
A good way to eliminate options is go one by one through the options and ask if anyone would be strongly opposed if that one ended up being the decision reached through the process. If anyone says they would be, ask them why. Based on what is said, as a group decide whether or not to eliminate the option from consideration.
2. Agree on Criteria
Similarly, everyone should have reviewed the criteria before the meeting. Spend 30 minutes or less gaining agreement on the criteria. Start by reviewing the filters and gain agreement that these are the general themes that will determine the right decision. Then ask if anyone has any concerns with the individual criteria within the filters. You don’t need to walk through each of the criteria.
This is a good time for me to emphasize that the discussions that happen during the sifting process are much more valuable than getting a precise score for each option. You don’t need to have a perfect list of criteria. You do need to be talking about the right topics — so it is essential that you have the right filters, but it is not essential that you have all the right criteria.
Keep the list of criteria within each filter down to less than 15 and preferably less than 10. (You’ll thank me for this later 😀.) Similarly, keep the number of filters to 5 or fewer.
Often a participant will make the point that not all of the criteria are equally important and will argue for prioritizing or weighting the different criteria. Resist the temptation. That will add a lot of work and not meaningfully change the outcome. Ask if any of the criteria are unimportant enough to remove. If not, then simply move on. The point of the criteria is to make sure you are thinking about and talking about the right things.
3. Walk Through Filters
You are now ready for the meat of the sifting process. This will take about an hour for the first filter and about 30 minutes for each subsequent filter.
You are going to walk row by row through the filter spreadsheet. For each criteria, you will ask the group to force rank the options against that criteria. The option that is most favorable gets the highest integer (if you have 4 options, that one gets the 4). The option that is least favorable gets a 1. Rank the options in between so that there’s one of each integer. Refer again to the sample spreadsheet, re-shown here:
Row 3 of this example says that the group agreed that the “Street Cred” option best leveraged the company’s “Mobile Event Check-in” capability and the “Online Campus” option least leveraged it. “Greek Guide” was second best and “Small & Virtual” was third best.
Disagreements are where the value of the process really shines. When one participant thinks a given option is the best and another participant disagrees, it is essential for participant #2 to speak up: “Why do you think that, I disagree.” Participant #1’s response will likely help participant #2 (and others) see that option or that capability (in this example) in a different light. They will probably learn about the company, it’s options, and each other in ways that will pay dividends well beyond the decision being made.
Often, someone will say that there’s no difference between two options on a given criteria. Unless that’s literally true (the two options are absolutely identical as it relates to that criteria), force the team to rank one higher than the other. If they are literally identical, split the difference. For example if they tie as the second worst option, give them each a 2.5.
Other than when options are literally identical, resist the temptation to use fractions. Mathematically minded people (e.g. engineers and accountants) will argue for relative scoring (e.g. 9.5, 9.1, 8.5, and 2 on a relative scale rather than 4,3,2,1 with forced ranking). Don’t do it. It likely won’t change the outcome but instead will push the different options together in score, making it less clear what the right answer is. The point of the process is to uncover how the options are different from each other and to spend time discussing that, not on debating whether an option gets 9.1 or 9.2.
4. Consider Results
Once you have completed walking through the filters, turn to the Summary page. Here’s the sample one again:
If you are using the template that I’ve linked below in the “Related Resources” section, you’ll notice that this spreadsheet has two sections comparing the options against the filters. The top half of the sheet uses the average scores from the filters (e.g. row 11 in our sample filter) while the bottom half uses the ranking of the options in each filter (e.g. row 12 in our sample filter).
Always start with the bottom half. Mathematically, the best option will be the one with the highest rank. Sometimes people will react poorly to this “right answer”. If the process has involved a reasonable amount of discussion and debate, participants will be emotionally engaged in the process and it’s likely that the “best option” has already become clear. When the math points to a different option, then they respond emotionally. That is very healthy and helpful as it leads into a discussion about why the highest ranked option isn’t the right one — it helps the group remember and reflect on what they have learned through the process and that will help them as they and their teams participate in implementing the truly best option.
Most of the time, the math actually works out so that the best option gets the highest ranking. Take the time to talk about why it is the best option to get that same kind of reflection that will serve the organization well going forward.
In the example given, two options actually tied with the highest ranking. That too is an opportunity to reflect and discuss. In this example, I would quickly look at the average scores and see that the “Greek Guide” option scored just slightly higher than the “Online Campus” option. More importantly, the “Online Campus” option scored very high in 3 out of the 5 filters, but scored very poorly on leveraging existing capabilities. “Greek Guide” scored high on leveraging capabilities and creating company value, but was just middle of the road on the other three filters. The group should discuss how those observations will impact your final decision.
No matter what the math says, it’s important to spend 30 minutes to an hour talking about the options that are in the final consideration. Look at how they did against each of the filters. From what you’ve learned through the process, what will be the greatest challenges in implementing this particular decision? What are the strengths of an option that you can leverage in building support and momentum behind it?
Before I leave this step, I do want to come back to the argument for forced ranking. The average scores shown in the top half of my example are similar to what you might get in each filter if you use relative scoring. There’s only 12% separation between the highest scoring option (Greek Guide at 13.6) and the third highest scoring (Small & Virtual at 12.1). Someone passionate about the Small & Virtual option could argue that difference is too small to justify eliminating the option. However, in the bottom half, the difference between those two increases to 50% (15 vs. 10) making it much easier to focus the final decision on the two options that truly stood apart.
5. Make the Decision
The point of the Strategy Sieve process isn’t to force the organization to adopt the option that gets the highest score, but rather to deeply explore and learn through the sifting process. The most common outcome I’ve seen using this tool is that the right answer becomes clear (whether or not it’s the highest scoring), everyone get’s on board with the decision, and all participants are ready to explain to anyone why it is the right decision.
But there are two other outcomes worth discussing.
Sometimes the process uncovers some important missing information. Some people think “A” is true (e.g. “customers truly want an automated solution”). If that is true, then the right decision is clear. If it is not true, then a different option becomes the right decision. Agree to test that hypothesis before making a final decision.
Other times, one option seems to be the right decision, but there are some issues that represent significant risk to its success. There are elements of another option that would address those shortcomings. In this case, is there a way to adopt a hybrid option that combines the best of both into a right decision that everyone is more confident will be more successful? (Be careful when pursuing this approach, it can open the door to participants reintroducing their favorite options and mushing all the options together into something that would be unworkable and a disastrous decision.)
Over the years, I have found the Strategy Sieve to be a very helpful tool in making hard decisions. I hope it can be equally helpful for you.
Related Resources
Hopefully this article has helped you better understand the Strategy Sieve and how to use it. Let me know if you ever need any help. You might also check out the resources below for more information and examples:
- An Example Google Sheets Strategy Sieve Workbook (save a copy and modify for your own use)
- A video tutorial on using the Strategy Sieve
- What is a Strategy Lab?
- Whither the Wind Blows (Impulse Drive’s use of the Strategy Sieve)
- SDG Games’ Revenue Model (Use of the Strategy Sieve for SDG Games)
- The Strategy Sieve
