Staffing Plan: XXX Team at XXX Larry Davis, Manager Some words have been changed to protect the identity of the sources for which this report was written and these changes have introduced formatting errors. Please excuse these errors as artifacts necessary to allow me to use this document as a writing sample. Directives 1. Reduce staff. 2. Ensure ability to provide coverage during stated operational hours. 3. Ensure ability to pass monthly SLR’s by a slim margin. Current staffing model The Enterprise Service Desk currently has a staff of 16. The positions are broken down into two main classes “Helpdesk I” and “Helpdesk II” with pay rates being higher for the “Helpdesk II” slot. Our staff is made up of 3 XXXX employees and 13 subcontractors of which 11 are from XXXX and 2 are from XXXX. Current staffing model. Company Number of staff HD Support I HD Support II XXXX Titles XXXX 3 1 - Client FTS Analyst 1 - Client Tech Support Spec 1 - Client Tech Support Sr. Analyst XXXX 11 5 6 XXXX 2 2 Totals 16 7 6 3 Although the Enterprise Service Desk is allotted 16 analysts, it can be seen by the current breakdown of workable vs. non-working staff that we are at 56% of full staff and far from attaining our recommended service levels and in fact have never reached 75% of workable staff due to conditions revealed later in this report. Current staff available for scheduling
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Staffing Plan: XXX Team at XXX
Larry Davis, Manager
Some words have been changed to protect the identity of the sources for which this report was
written and these changes have introduced formatting errors. Please excuse these errors as
artifacts necessary to allow me to use this document as a writing sample.
Directives
1. Reduce staff.
2. Ensure ability to provide coverage during stated operational hours.
3. Ensure ability to pass monthly SLR’s by a slim margin.
Current staffing model
The Enterprise Service Desk currently has a staff of 16. The positions are broken down into two
main classes “Helpdesk I” and “Helpdesk II” with pay rates being higher for the “Helpdesk II”
slot. Our staff is made up of 3 XXXX employees and 13 subcontractors of which 11 are from
XXXX and 2 are from XXXX.
Current staffing model.
Company Number of staff
HD Support I
HD Support II
XXXX Titles
XXXX 3 1 - Client FTS Analyst 1 - Client Tech Support Spec 1 - Client Tech Support Sr. Analyst
XXXX 11 5 6
XXXX 2 2
Totals 16 7 6 3
Although the Enterprise Service Desk is allotted 16 analysts, it can be seen by the current
breakdown of workable vs. non-working staff that we are at 56% of full staff and far from
attaining our recommended service levels and in fact have never reached 75% of workable staff
due to conditions revealed later in this report.
Current staff available for scheduling
Hours of Operation
The Enterprise Service Desk utilizes a modified 15 x 7 dedicated support hours of operation model, with 15 x 5 Monday through Friday and 12 x 2 on weekends. Current hours are set to exceed normal business hours and are from 6am-9pm Monday through Friday and 9am-9pm on weekends.
Enterprise Service Desk has been unable to supply support for weekend coverage up to this date. The cost in man-hours to cover such low volume need on weekends has not supported the loss of manpower through the week that we would lose in order to cover these hours with existing staff. Until such time as the Enterprise Service Desk gets enough cleared personnel to afford to cover the weekends, these shifts will continue to be covered by existing NOC staff. Current data suggests we take an average of 1 call every 3 hours during the operating hours of 9am-9pm on weekends.
Man-hours needed to conduct operations
The Enterprise Service Desk handled an average of 5950 monthly contacts over the last two
calendar years (2011 & 2012). These contacts are broken down into four main contact types,
Telephone, Email, Web based (Service Catalog) and Walk ups of which telephonic has the lion
share with 70% of the total contacts on average over the last two years (See Graph: Two year
average of contact types). The workload of the Enterprise Service Desk can be seen to be
relatively stable with an increase during the first quarter of the year that declines and remains
steady until year end (See Graph: Monthly total of all contacts for two previous years). Of the
contact types we record, the percentage of each type also remains relatively constant with some
variations due to inadequate recording during the process (See Graph: Annual contacts by type
with average of combined years in yellow) . The data is beneficial to as it is constant and so our
staffing needs can be seen to be just as constant and allow us to plan accordingly.
Two year average of contact types
Monthly total of all contacts for two previous years.
Annual contacts by type with average of combined years in yellow.
Hard data aside, we must make assumptions that allow us to adequately staff the Enterprise
Service Desk with sufficient manpower to meet our SLR’s and supply satisfactory service to the
XXXX. We can prove that the Enterprise Service Desk handles about 6000 contacts per month.
We know that our operating hours are 15 per day per work week, and 12 per day on the
weekend for a total of 99 hours per week. Breaking that down simply, the call volume per hour
equates to about 60 calls per hour or 1 call a minute. Fortunately for us we have better records
and can almost eliminate the 24 weekend hours and the week-day hours from 6pm-9pm from
our calculations as the number of calls we get during these times is negligible and actually
serves to throw off the calculations. With 60 hours remaining, our “calls per minute” stat shows
an average of 3. Data shows (See Table Average Call Time) that an average call lasts 6
minutes. During that first set of three six minute calls the Enterprise Service Desk will have
gotten 18 additional calls, each needing a free analyst to answer and each answer requiring 6
minutes of the analyst’s time. When broken down using these simple metrics we are faced with
real data that shows we need 18 analysts in order to cover that workload, all on the phones and
all within a six minute span of time. Although this certainly occurs, it does not occur every
minute of the work day as the data suggests, but instead comes in peaks and valleys that can
be scheduled around (for the most part) by graphing the average calls per hour over time (See
Graph: Hourly Call Volume). In order to make the data more meaningful I have broken it down
into calls by hour for a one month span of time split into four weeks by hour of the day and
averaged a “calls per hour of the day” metric that is accurate for the Enterprise Service Desk
workload and adjustable as new service offerings come online.
Table: Average Call Time
Taking this data I have charted the number of analysts needed to handle hourly call volume
based on the metric that one analyst can handle between 0-10 calls per hour on average (The
real average allowing for 1 minute work time between calls is 8.57). We can expand this up to
about 15 calls per hour with some loses to our SLR’s and possible failures if sustained for too
long a period. Call volume of 16 calls per analyst and above almost guarantees multiple SLR
failures based on current data. The table “Current division of work Mon-Fri “ shows real hourly
call volume (averaged) against our current schedule with a color coded “Calls per tech” row
based on the information provided in this report. For five of the fifteen daily hours we are
exceeding our limits and failing SLR’s without even taking into account the daily call outs and
late arrivals. This outcome is supported by the four SLR failures reported for February.
Legend for Table: Current division of work Mon-Fri
Avg Call Time Calls Taken
Bill 0:08:49 388
Dillon 0:05:01 464
Donna 0:06:28 345
James 0:05:27 418
Jude 0:06:09 273
Maurice 0:05:43 154
Ryan 0:06:41 473
Tasha 0:02:47 348
Tom 0:07:34 325
Average 0:06:04 354
Calls Code
0 - 5 Blue
6 - 10 Green
11 - 15 Yellow
16 - 20 Red
21 - 30 SURGE
Table: Current division of work Mon-Fri
With the addition of one man to our current workforce scheduled to cover the peak hours and
allow for more flexibility and a relaxation on overtime bans for subcontractors I think I could
provide the service we need in order to ensure we are almost all green. The final consideration
would be that I need employee’s “In the pipe” that are on the way to approval so that losses can
be recovered quickly. With this tight of a line, there is no room to work without a ready spare to
replace the flat. The table below has the additional man and shows how we are now almost
completely within reason of reaching current SLR targets. This proposition sits on a razors
edge, needing only a very little to push us beyond our ability to meet targets. The addition of a
second man to this table would drop the yellows from being almost red to almost green, still
meeting the target of just passing SLR’s, but perhaps at a more realistic rate (Table not shown).
SLR 11.a Call Answer Time Average < 45 seconds for 95% of calls Cisco ACD
SLR 11.b First Call ResolXXXXon 85% of calls must be completed on that call Remedy
SLR 11.c Call Abandonment Rate (DISPUTED) 85% of all calls must reach an agent
Cisco ACD
SLR 11.d Average Call Hold Time Average of <91 seconds for 95% of calls Cisco ACD
SLR 11.f Email Response < 8 business hour response for 95% of emails Remedy
SLR 11.g Web Portal Response < 2 business hour response for 95% of emails Remedy
SLR 11.h RoXXXXng of Incidents 95% of “incidents” (No SLR for Requests) correctly routed
Remedy
SLR 11.i Availability of Live Agent Available for 98% of service hours Cisco ACD
Risks
1. SLR’s and the SOW are poorly written and subject to change with little or no notice once
they are reinterpreted by the XXXX. These changes can have a direct impact on the
staff needed to meet the new interpretation.
a. Recent examples of SOW re-interpretation
i. The operational hours of Desk Side were reinterpreted to require staffing
until 9pm when it was previously accepted to be 5pm.
ii. The SLR of the Service Desk concerning average hold time was re-
interpreted to require a “per call hold time” that caused all of our
submitted data with monthly averages to be rejected and the Enterprise
Service Desk to fail this SLR for months before we were able to provide
the data according to the new interpretation.
b. XXXX is currently reviewing the formula that dictates the number of abandoned
calls the Enterprise Service Desk is allowed by volume. If passed, our current
staffing model will be insufficient to allow us to continue to pass this SLR.
c. The XXXX is aware of mistakenly omitting weekend SLR’s and intends to correct
that at some future time.
d. Additional service years sometimes carry stiffer SLR’s as is the case with the
Enterprise Service Desk’s First Call Resolution SLR that went from a high
industry standard of 65% to an unheard of 85% completion target.
2. The model of the Service Desk is being questioned. If our model changes due to
continued applied pressure, it could have a great impact on staffing levels that we are
unable to forecast here.
3. During contract start XXXX reported staffing levels across all operational departments.
These numbers were submitted formally as ORR’s to the XXXX. This could present a
challenge when we reduce the numbers that have been submitted and are a part of the
XXXX public records.
a. Sub-contractors who were also guaranteed a number of slots may also protest
the reductions.
4. Running with a staff model that only adequately passes and sometimes fails SLR’s will
not be adequate for Operational surges and new services. We could encounter a sine
wave effect of sudden failures for some months until we are able to staff up to meet the
new demand only to fail during the next new service offering.
5. Overtime control will become problematic as we attempt to fill for vacations and other
time off but do not have staff to cover.
6. Event management may be impacted as the XXXX desires members of the staff to
support local events such as work at home day, the RIK or the recent Mobility event in
which two Enterprise Service Desk staff have been requested for.
7. Shortages due to turnover have been buffered by the staff we have “in the pipeline”.
With a reduction or elimination of the staff waiting to be cleared during the lengthy
badging process we will run short, and if the past is an indication we may not have the
support staff needed to operate stated business hours. Without those numbers who
were already in the clearance process we would have faced many occasions where we
did not have the operational staff to support our operational hours.
a. The Enterprise Service Desk has never had sufficient cleared staff to allow
weekend operations with current guidelines; a further reduction will incur further
shortages.
8. The Enterprise Service Desk has 8 SLR’s directly related to work confined to the Service
Desk but our work indirectly impacts almost every other service related SLR that is
submitted via the Enterprise Service Desk in ways that could easily cause other teams to
fail SLR’s due to shortages or overload on our end. As an example, if a VIP emails the
Service Desk for an Email Restore request (SLR: 2 hours) and due to call volume the
desk cannot spare manpower for email management, the delay in creating that ticket
could cause the Email & Messaging team to fail that SLR for the month. Due to the
infrequency of these requests it has only taken one failure for the month to fail the
associated SLR.
Recommendations
1. Combine the three operational teams that make up the Service Desk (NOC, Enterprise Service Desk, Desk side) to achieve our goals without sacrificing service levels. The
XXXX model of the service desk includes Tiers 1 through 3, 1 being the Enterprise Service Desk, 2 Desk Side and 3 the NOC. We already have the NOC on a standing 24x7 schedule and can use that resource as our backup if the main Enterprise Service Desk analyst does not show for weekend duty.
a. This would negate the need to staff at twice the needed capacity just to ensure we do not go dark if we get a failure to appear over the weekend and eliminate 24 weekly hours.
i. Currently we are letting the XXXX tell us that the NOC is not part of the Service Desk, even though by XXXX design it is. The XXXX reports that they want someone actually sitting in the Service Desk area, but in their own statement of work they allow that at some future point we may have an off-site service desk (SOW 33-1--1325: There is not a requirement for service desk staff to be co-located with XXXX staff beyond the need to maintain sufficient staff at local sites for providing deskside support) making that argument moot. If we pressed this further, we could eliminate the need for a “Enterprise Service Desk” analyst on site weekends altogether.
2. Close desk side support at 6pm as in the past and add a 2nd Tier 1 worker in the
Enterprise Service Desk from 6pm-9pm. The 2nd Enterprise Service Desk worker would
be used as needed to dispatch to desk side.
a. This further blurs the line between the teams showing the XXXX that we are one
team, and are all able to respond and interface with customers as needed. As
the Enterprise Service Desk has a lower pay scale we have a cost savings.
b. As everyone gets accustomed to this change we could go one step further and
replace the additional 6pm-9pm worker with an existing NOC member who would
be assigned to answer the phones while our Enterprise Service Desk worker
responds to desk side as needed.
c. This change could be a mid-term gain of a reduction of 15 weekly hours.
3. Place the responsibility of obtaining a clearance prior to hire on the hiring authority.
Months of usable work is lost on every new hire as they sit almost idle at the PMO
unable to perform the true functions they were hired for. We could make having a
clearance part of the requirement prior to hire.
a. This may not be possible to enforce at XXXX, but could be applied to our
subcontractors and in fact was applied early at the start of the contract. The
subcontractor would select candidates who are currently employed but looking to
move and conduct the clearance process while they are still actively employed
elsewhere.
b. This would reduce months of wages provided without return and end the situation
of having waited some months just to find that the candidate has been declined
or found another position before he even provided one real days work.
c. The gains of achieving this change are beyond my ability to calculate, but should
be at least a 20% savings in wages or the equivalent of cutting three analysts
from the team without a reduction in staff.
4. Increase the wages and benefits of existing positions.
a. According to major staffing studies the current XXXX wage for Enterprise Service
Desk workers is fully 30% below this areas average, and due to the team being
staffed by contractors they also see almost no benefits associated with that lower
than average wage.
b. By institXXXXng an immediate increase in wages and benefits we would
dramatically reduce turnover and at the same time attract a more qualified
candidate. It can be assumed that these new candidates will be able to produce
at a higher level and reduce the need for more staff on the phones by reducing
the average time on a call through increased competency.
i. The disadvantage with this process is that it will not happen quickly. The
higher quality candidates will have to be brought on through attrition,
obtain clearance and then trained into the specifics of this environment.
Accounting for poor hires that do not produce even though they seemed
capable during the hire process and other factors will make this change a
long term substantial win that bears no short term fruit.
5. Abandon the Subcontractor model of the Enterprise Service Desk.
a. Bringing the staff in as XXXX employees would attract a better class of employee
in the same way that a wage and benefit increase would.
b. This would also eliminate the extra overhead that is needed to manage the
different sub-contractor accounts and time accounting.
c. The current ban on sub-contractor overtime would become a non-issue allowing
for better shift control and less fluctuation due to shortages that cannot be
replaced without incurring that overtime.
d. Having XXXX employees will ease the transition from one job to the next as we
attempt to hire from within, something that is extremely complex if not impossible
currently.
e. Combining this with a wage better suited to this area and XXXX’s benefits should
enable us to compete for the best talent available and reduce overhead as we
bring the team up to the standards these changes allow.
6. Reduce headcount
a. Depending upon need and urgency a direct reduction in headcount before or
regardless of any of the above recommendations is put into effect would be
possible in the short term.
b. Historically, XXXX has been unable to reach a level of support sufficient to both
meet our SLR’s and staff the weekend schedule. The sixteen slots allotted to the
Enterprise Service Desk have never all been filled at the same time, and the
most ever cleared at the same time has been less than 10.
c. A reduction in headcount would not immediately been felt by the Enterprise
Service Desk as the team is used to losing and then getting someone at the
PMO cleared to take the place soon after. The loss would be felt when the
pipeline runs dry due to the reduction and the short term loses become
permanent, or worse yet, dip into losses not yet experienced by this team.
d. We have had occasions when we were just able to maintain operations due to
staff shortages, and this with 16 available positions, what will our service
offerings look like with less? Time will tell, but without remedial actions on the
back end to reduce the need for these slots, the results would not be acceptable
at any level.
7. Reduce the number of contacts
a. XXXX currently has no dedicated user training program and yet almost every
system we have carries some type of training requirement. We should institute a
procedure that requires service owners to provide user based continual training
on the use of their service offerings in order to reduce the number of calls that
come into the Enterprise Service Desk.
i. XXXX could hire a dedicated trainer to handle this for services across the
board if the burden of training would overtax service owners.
b. Through continual process improvement. Identify the type of contacts we receive
and train / improve around those most frequent calls in order to reduce them.
i. The Enterprise Service Desk is already putting out stats that show our top
calls and works with other teams on ways to reduce them, but with the
amount of change XXXX is making to this environment, progress has
been slow to occur as the various service teams devote their resources to
current fires.
c. Increase the usefulness of the Self Service portal by making it more user friendly
and providing a means to get self service for most of our top ten service
requests.
d. Advertise the self service and training offerings we make available through the
XXXX front page and various announcements or “Open house” campaigns to
increase awareness.
STAFFING PLAN
Based on the data and recommendations I’ve provided, I propose;
1. An immediate reduction in staff by two slots. This reduction is already in the works as I
have two open XXXX positions that contracts is looking to remove.
2. Seek to bring as many of the 11 remaining sub-contractors in-house with a wage
suitable to this areas standard for the type of hire we seek (a Tier II analyst with
telephonic and desk side experience).
3. Increase the wage and benefit package for the subcontractors that remain.
Once these changes are made we report on performance every three months.
During the first quarter we should see a more steady SLR rating as shortages due to call offs
and overtime are reduced. We will not see an improvement due to the increases beyond
increased morale as there will not have been sufficient time to interview, hire and clear anyone
brought in with the new wage/benefit structure. Further I would not rehire the next Helpdesk I
analyst that leaves our employ.
During the second quarter we may have incurred our first loss, but as we are not going to begin
up-staffing until another loss, we will not be able to identify and quantify the level of competence
our changes have increased for some time. Bringing in just one of these “senior” analysts could
help the entire team as they have another “go to” guy to hand off the harder issues to.
It’s in the third quarter that we may be able to determine the extent of the second reduction in
head count. Are we exactly where we were when the changes were made, or have we
accomplished what we set out to do? Small changes will provide small results, but if we made a
difference that has attracted the best talent and reduced our turnover my goal would be to
further reduce the headcount by 1 by the end of this quarter and maintain a working team of 9
with 3 in the pipe at all times, provided we can get sub-contractors to agree to not hiring until
cleared.
CONCLUSION
The current staffing model for the Enterprise Service Desk is not working. The limitations
imposed by using sub-contractors along with a sub-average wage and almost no benefits
ensures high turnover within a system that has built-in lengthy delays between hire and
inception. This system has not allowed us to fully staff our operational hours despite almost
constantly processing and onboarding of replacement workers and is causing massive waste as
we pay for un-cleared workers to sit at the PMO for months only to recoup a very limited return
as they quit for a better job shortly after (and sometimes before) clearance is obtained.
We should retool our model and channel the money we spend on maintaining the revolving door
into a better class of worker who will work for the duration, be promotable and provide superior
service, reducing the number of workers needed to perform the same amount of work. It’s a
known fact that hiring and training is expensive, add the months of non-work on top of that and
it’s clear that we must focus our efforts on attracting the best, and keeping the best.
We should also investigate the possibility of requiring, or at least strongly suggesting that
remaining subs focus their hiring efforts on candidates that are currently gainfully employed but
what to move up or change employers so that we are not paying them during the months long
clearance process, and are not paying them at all if they fail.
This will not allow for the high front end reduction that may be desired but will be an investment
in the future that will ensure a larger pay off over time with the benefit of more stable SLR
numbers and increase service levels across the board.