Category list

Preparing an ROI Justification

Many CRM purchases include the ritual of a Return On Investment (ROI) analysis. A ROI analysis contrasts the costs of the CRM project with the expected benefits, and among other things forecasts how long it will take before the benefits outweigh the costs. ROI analyses traditionally focus on the cost savings created by the increased productivity made possible by the tool, but it's also important to consider potential revenue increases the tool may bring about.

ROI are often required to get the purchase approved. While many companies used to take a leap of faith into CRM without making a quantitative justification, this is changing as top management is insisting on cost justification. Even if you don't have a corporate requirement to prepare a ROI analysis, it's very useful to map out what you will get for your investment. You may even find that you curb some of your desires for fancy customizations when you realize that the payback is not commensurate with the expense.

While a thorough ROI analysis is the domain of professional financial analysts, the impetus for the ROI and much of the benefits estimates must come from the business owners , so they need to be the ones driving and validating the ROI process.

This section contains checklists to help you organize the ROI analysis and some examples to get you started.

Costs

It should not be difficult to define the costs (there's much less guessing involved compared to the benefits!) but you need to be disciplined and include all costs, even indirect ones:

License fee for the tool.
Maintenance and support fees . If you are doing a multi-year ROI analysis (and most CRM tools require several years to break even), be sure to take into account likely increases in the fees.
Purchase costs for ancillary items , including hardware and software. It's not unusual to see ancillary costs approach the costs of the tool.
Maintenance and support fees for the ancillary items .
Implementation costs . They could run much higher than the license cost, depending on complexity.
Training . Include end- user training and training for developers and the system administrator. Costs could be out-of-pocket (to the vendor or a third party) and/or internal costs, especially for end-user training.
Internal staff . Internal costs are rarely included, and part of the reason is that they are difficult to compute, especially for staffers that only spend part of their time on the project. I feel it's important to capture at least the cost of resources that are spending a significant percentage of their time on the project including technical staffers , the project manager if that's an employee, and super-users .

Table 7.2 shows an example of a cost analysis for a modestly sized initiative that includes internal costs. Note that, although the implementation is farmed out to an integrator, there are significant IT costs to support and maintain the tool. Further note that, without the internal costs, the Year 1 costs would be only 65% of the total shown here, so internal costs are very significant.

Table 7.2. Sample CRM Project Cost Analysis

Item	Year 1	Year 2	Year 3
License	150	^[1]
Maintenance ^[2]	30	32	35
Hardware cost	40
Hardware maintenance ^[3]	7	8	8
Database system	10
Maintenance ^[4]	2	2	2
Implementation	120
Training/IT	35	10	10
Training/end-users ^[5]	25	5	5
IT staff cost	120 ^[6]	100 ^[7]	108
Super-users cost	60
Business owners cost	40
Total	639	157	168

^[1] This assumes no additional purchases, which may or may not be realistic for you.

^[2] This uses a 20% of license price and a yearly increase of 8%.

^[3] This figure is based on an 18% price structure and a yearly increase of 8%.

^[4] This assumes a price structure of 20% of license and a yearly increase of 8%.

^[5] The training figures assume no travel. Training for years 2 and 3 is for updates (IT) and new-hire training (end-users) respectively.

^[6] This amount is for the selection and implementation project and the support and maintenance after the rollout.

^[7] This amount is for the support and maintenance of the tool.

Benefits

Now that you have a big fat cost number, it's time to add up the benefits to see which side wins and how quickly.

Baseline

Before you can define savings, you need to establish a baseline for your current situation and costs. To that end, you need to establish the number of CRM users, broken down by categories, their compensation (averages by category are sufficient) and the key metrics for the tasks they accomplish. Key metrics for sales reps include the average deal size, the closing ratio, the number of appointments or proposals per month, etc. For service reps, this would be the number of issues resolved per month, the ratio of reps to customers, etc. Chapter 10 of this book explores the topic of metrics in depth, but at this point you're only looking at basic metrics that capture your cost of doing business. They should give you the current cost per sale and the cost per support case, as well as a structure to forecast the impact of the new tool.

It may be difficult to gather reliable data about the current situation, and indeed this may be the impetus for getting a new tool in the first place. Nevertheless, you must estimate the missing data as best you can in order to make progress on the ROI analysis.

Cost Savings

Once current costs are known or properly estimated, then you need to estimate the productivity improvements brought about by the new tool. If today service reps close 10 cases a day, will they be able to close 11? More? If sales reps close (win) 70% of the deals, can they win 80%?

Estimating productivity improvements is difficult. The best approach is to break down the tasks being automated by the project both by specific individuals and specific areas of their jobs, and to estimate the savings on each subtask rather than making a global estimate. If sales reps spend 25% of their time preparing for sales calls and the tool allows a productivity improvement of 20% on that activity, perhaps by making it easier to gather relevant materials or by improving collaboration amongst the sales team, then the overall improvement for sales reps' productivity is 5% for that particular activity. (Note that a 20% improvement is very large indeed, so you should question and revisit your assumptions if you should come across such large numbers ).

Table 7.3 shows an example of how you can use a subtask analysis to forecast the efficiency improvements the tool will bring. Note that the end figure, 8%, is not so large as to be suspiciously over-optimistic.

Table 7.3. Example of Productivity Improvements for a Sales Group

Activity	% Of time spent	Projected % Improvement by Activity	Overall Impact %
Making appointments	5	2	^[1]
Pre-call planning	25	20	5
Sales call	20	^[2]
Proposal creation	5	10	1
Contract negotiation	10
Travel	10	5 ^[3]	1
Follow-up work	10	5	1
Reporting	5	15	1
Administrative tasks	10	5	1
Total	100%		8%

^[1] This is not really 0% (it's really .1%).

^[2] It's rare that a particular area would show no improvement whatsoever, but don't be afraid of admitting where the tool will not help.

^[3] Can a tool increase the effectiveness of travel? Yes, if it helps prevent unnecessary travel.

To give you another way to approach the same problem, Table 7.4 is an example for a support group that focuses on how the tool will expedite the resolution of an average support case. Here again, the forecasted efficiency gain (about 10%) is reasonably small and believable, although you would need to dig a lot deeper to validate each of the assumptions.

Table 7.4. Example of Productivity Improvements for a Support Group

Activity	Minutes Spent Now	Projected % Improvement by Activity	Forecasted Minutes Spent
Check entitlement	1	20	0.8
Log case	3	5	2.85
Dispatch case	1	^[1]	1
Resolve issue	15	20	12
Escalate to level 2 ^[2]	1		1
Level 2 resolution	10	10	9
Wrap-up	1	10	0.9
Total	32		28

^[1] Estimating improvements for tiny steps is usually not worth it, although if this were a very large support center with thousands of calls it would be worthwhile to do so since everything adds up. For instance, CTI typically shaves just a few seconds off a case's processing, but shows clear overall savings for heavy-volume centers.

^[2] Not all cases are escalated to Level 2 so the numbers here reflect the escalation effort for the average case, not the actual escalation effort for cases that are escalated.

The business owners should be very careful about creating the list of tasks and their distribution since the entire analysis will be driven by the numbers. For instance, in the support example above, if the Level 1 resolution actually takes 30 minutes rather than 15, the overall efficiency improvement in case handling time would be a whopping 16% ”from 47 to 40 minutes ”compared to the already respectable 14% figure ”from 32 minutes down to 28. So make sure you get the big pieces right rather than obsessing over the details.

Here are some suggestions for how a CRM tool can improve efficiency for a sales environment. Use them to jog your memory as you analyze the subtasks .

Will the tool increase the productivity of telesales agents ? Will they be able to handle more conversations during a given day? Think both of outgoing and incoming contacts, as appropriate. You may want to break out each step of contacting customers if you have a large group or are implementing specific technologies for that team.
Will the tool help generate better quality leads, thereby improving the conversion ratio?
Will the tool improve the distribution and routing of leads?
Will the tool help sales reps prioritize and organize their work more effectively?
How much time do sales reps spend locating information? How can the tool decrease that? You should get nice savings here.
Will the tool help sales reps communicate with customers more effectively?
How much time do sales reps spend locating information? How can the tool decrease that? You should get nice savings here.
Will the tool decrease administrative chores for sales reps?
Will the tool decrease the need for sales calls? Will it help decrease the length of sales calls? Expedite their preparation? What effect if any will it have on travel required?
Will the tool speed up the generation and approval of proposals and quotes? This is particularly important if you are operating in a price-sensitive environment.
Will the tool allow for more accurate sales orders? This is a key benefit if your product is complex and you are introducing a configurator. Inaccurate sales orders in turn could create costly returns and changes.
Will the tool help increase deal size? This may be through decreased discounting (through better management control perhaps), or the ability to call higher into accounts (through better enforcement of sales methodology), or simply access to better information.
Will the tool increase the win ratio?
Will the sales cycles decrease? By how much?
Will the tool allow sales reps to spend less time on post-sales activities such as checking on delivery or service?
Will the tool allow sales managers to recruit, train, and manage sales reps?
Will the tool help reps and mangers to create forecasts?

Here are some analysis questions for support centers.

Will the new tool decrease the volume of incoming issues? Fewer issues coming in mean less work. This is a big deal for self-service initiatives.
Will the new tool expedite customer entitlement and logging of issues? This may be significant if customers will now be logging their own issues electronically .
Will the new tool save time in routing issues? For instance, a new CTI system can deliver questions to the right agent automatically. Even tiny savings here can add up if you have a lot of incoming calls.
Will the tool save time during issue resolution? You may want to break down the tasks further here and consider easy, repetitive issues separately from complex issues. For easy issues, you should save big with a new knowledge base, especially if you can resolve issues in one interaction. For complex issues, you may save time with automatic escalations or collaboration features. Can the tool prevent escalations to Level 2 entirely?

The tool should make it easier to communicate with customers, so you can shave some time there too, and especially for complex issues.
Will the tool save time administering the user database? For instance, self-registration on the portal may be a big help.
Will the tool save time in collecting and sharing information? A cumbersome document creation workflow may be cut down to size with a good system for writing, reviewing, and posting documents.
Will the tool save training time and resources? This can be significant if you have high turnover , or lots of new products and systems to train on.
Will the tool save time in obtaining appropriate metrics for the managers? And will the very existence of the metrics allow for savings or increased productivity? For instance, a workforce management tool will shine here.
Will the tool make for easier revenue collection, whether you sell yearly contracts or incidents?
You may be able to find additional inspiration in Chapter 10, which discusses metrics.

Once you have estimated the productivity improvements, combine them with the cost data to figure out cost savings. This can be a pretty complex operation, perhaps better suited to the talents of a financial analyst, but the idea is simple. If today you accomplish a given amount of work (sell a million dollars' worth of goods or support 1000 customers) with X resources costing Y, with the tool you would be able to accomplish the same amount of work with fewer than X resources at a cost that is less than Y.

Revenue and Strategic Improvements

Most ROI analyses stop right here with the productivity improvements, even though productivity improvements alone rarely launch a CRM project in the first place. Instead, CRM projects often arise from strategic needs to meet market demands, to respond better to existing customers, or to reach new markets. Although it's harder to forecast or even to quantify some of these benefits, I feel it's important to at least give it a try. Here are some questions to answer to identify and quantify top-line benefits.

Will the tool allow you to deliver entirely new products, new services, or to reach new markets? For instance, you may be able to start an e-commerce channel, or to provide (and sell) proactive monitoring and alerts as a support option. Perhaps the tool will allow penetrating larger accounts. What is the size of the new market?
Will you be able to cross-sell more effectively to your existing customers? For instance, you may be able to use marketing automation to identify products your existing customers may be interested in. Or you may increase your support contract renewal rate. What additional revenue can you expect from cross selling?
What will be the impact on customer satisfaction? It may be negative, for instance if you are forcing customers to use self-service, but on the other hand customers may be clamoring for a level of service you cannot deliver without the tool. Transforming customer satisfaction improvements into dollars is an art in itself, so try something tangible such as increasing the retention rate.
Will the tool increase employee satisfaction? If the tool decreases turnover, perhaps as employees are able to spend less time on boring issues, you may be able to identify a significant upside.
What will be the impact of better analytics? This is a large area, starting with better revenue forecasting and pipeline analysis, useful to both the finance group and the manufacturing group. Better analytics also allow you to identify the top customers so they can get special treatment. They make it easier to cross-sell and up-sell existing customers, and to achieve better market penetration by replicating wins and avoiding repeated failures. Can you quantify the benefits of better analytics?

I like to be conservative in ROI analyses, keeping only tangible, measurable benefits and leaving the intangibles out of the numerical analysis altogether while listing them alongside the numbers. I feel that making sweeping assumptions about the intangibles suggests that the rest of the analysis may not be that realistic either and many CFOs agree with me. If you can't make your case without the intangibles, you may simply not have one.

As you prepare the benefits part of the ROI analysis you may find that, in order to gain some of the benefits afforded by the tool, you will need to make changes to your current business processes. To take a simple example, you may see that you need higher caliber sales reps in order to sustain the penetration of larger accounts that is made possible by the tool. If that's the case, you need to go back to the cost section and record the higher cost so you end up with a correct analysis, which is a good segue for the next point.

Are ROI Analyses Realistic?

The honest answer is: not very. Most ROI analyses are slapped together to create an acceptable justification, and as such occasionally massage the truth to show a neat, short- term benefit.

If you want to be able to stand by your analysis for the long term, be realistic rather than optimistic. In particular, sweeping productivity increases are rare. If you are forecasting a 20% increase in sales rep productivity, double-check your numbers. They may be right, but I bet you will find opportunities to temper your optimism .

Another area of raging optimism is how quickly benefits will accrue. In my experience, most projects experience a temporary decrease in productivity when the tool is rolled out, as users need to learn their way around the tool (and, not infrequently, technical problems need to be sorted out). Productivity improvements may not appear for weeks or even months after the system is rolled out. The more complex the system, the longer it will take for the improvements to show up.

Finally, ROI analyses are most accurate and useful when they focus on reasonably short-term wins, not fuzzy long-term chimeras. Make sure that your short-term benefits are accurately researched and justified. Then you can present the long-term benefits as icing on the cake.

The next chapter focuses on selecting an integrator. Integrator selection should occur at the same time as the final selection of the tool, so be sure to advance both paths concurrently.