Cloud Migration Deck

About this template

How much fits in 8 slides

An 8-slide deck for reporting migration scope, sequence, downtime and cost change in about 15 minutes. Server inventories, network designs and rollback runbooks do not fit; attach them separately and keep only what a decision needs.

Check the agenda against the slides first

The agenda has four items – asset inventory, before and after, key metrics, trend – numbered 01 / 04 through 04 / 04. Only the first of them has a divider-plus-body pair (slides 3-4); the other three are single slides: the comparison on 5, the metric cards on 6, the bars on 7. So the deck is one body section plus three summary screens. To show the migration wave by wave, duplicate the 3-4 pair, then renumber the agenda lines and the 01 / 04 markers to match.

If you add sections, use this order

1. Current state: system count, aging hardware, what gets retired
2. Approach: what moves as is versus what is rebuilt
3. Sequence: waves and the cutover window for each
4. Downtime and rollback: the trigger and who decides

Fix the numbers already in the file

Slide 4 carries 84 physical and 216 virtual servers, capacity and IOPS per database instance, and 47 linked batch jobs. Slide 5 sets before (42 servers in your own IDC, 4-hour recovery objective, 620M KRW a year) against after (managed services, 20-minute recovery objective, 380M KRW a year). Slide 6 holds 300 servers in scope, 4 hours of allowed downtime, a 7-month program and 31% cost saving. Note that 84 plus 216 equals the 300 on slide 6, while the 42 on slide 5 does not line up with either – so your first job is to settle on one asset count and carry it across all three slides.

Filling it in

Write the three body lines so scope, timing and method each get a line: “e.g. 12 read-only servers of 300 in wave 1”, “e.g. cutover Saturday 02:00-05:00”, “e.g. database moves by replication, 8 minutes of real downtime”. The bars on slide 7 read 22, 48, 76 and 112 for waves 1 to 4 under an index caption; replace them with a real unit such as servers moved per wave and rewrite the caption. The bars are shapes, so editing the number will not resize them.

Common mistakes

• Promising zero downtime before testing it. Slide 6 already fixes 4 hours of allowed downtime, so replace that with the window you have actually agreed.
• Leaving the stock before-and-after values on slide 5 (42 servers, 4 hours to 20 minutes, 620M to 380M KRW) as filler – they are the first numbers an approver questions.
• Showing savings while omitting dual-running and retraining costs.

Three questions the approver always asks

If the eight slides do not answer these, you will be back for another meeting. First, how long the service stops and when. Second, the last point at which you can roll back and who makes that call. Third, how different the first month’s bill will be. The first and third already have homes on slides 5 and 6, so you only need real values; the rollback trigger has no slide at all, so duplicate the 3-4 pair for it and add an agenda line.

Before you send it

A deck listing target systems and owners needs an explicit distribution scope. The cover has no presenter line, only a category label and a blurb, so put your team and the reporting date in that blurb, and check that the closing line on slide 8 is no longer the template default.