How to Analyze an eCommerce Replatform vs Redesign Decision
Build your own evaluation tool to fully understand short and long term implications of replatforming.
Shopify Plus, Magento Cloud, and a host of other new or (seemingly) improved eCommerce platforms promise to increase your revenue, decrease your technical effort needed and make your life a whole lot better — all if you use their platform.
I’m frequently asked what I think about upgrading (e.g., Magento 1 to Magento 2 or Cloud), re-platforming (e.g., Magento to Shopify) or just redesigning on an existing platform — and if it’s worth it.
Worth it is the key phrase here. To me, worth it means: “Am I going to get a return on my investment?”
Worth it analysis should be applied to every scenario, whether it’s a redesign, re-platform, or an upgrade. They each require an investment and provide their own unique pros, cons and risks.
Below, I break down how I typically answer this question, but first I need to outline key assumptions to my analysis to ensure we stay on the same page!
At a very simple view, increasing gross revenue from an eCommerce website boils down to:
- Increasing Traffic
- Improving Conversion Rate
- Increasing Average Order Value
Other metrics — like bounce rate, pages per session, abandoned cart %, etc. — are all secondary KPIs to the three above. See how I break down these three big metrics in more detail here.
Considering this is a theoretical analysis, I’ve invented some required data points to help us in this exercise. Let’s assume for this analysis, all these are true:
- Cost per hour for design and development work = $175
- Design and development effort = 800 hours
- Gross margin = 40% (to factor in the cost of goods sold)
- Total traffic = 1,200,000 sessions
- Average Order Value (AOV) = $500
- Baseline aggregate conversion rate = 0.50%
- We’re reviewing the scenario of 12 months of post launch.
- No additional offsite marketing initiatives are included as part of ROI calculation.
- No internal operational efficiency improvements are being included as part of our ROI calculation.
Data Set Definitions
In my visuals below, there are column headers to help organize the data and various ROI calculations. Before we jump in, here’s a quick explanation on what each of those mean:
- Estimated Conversion Rate Change: the % change from the baseline conversion rate; used to help calculate the …
- Actual (“New”) Conversion Rate: calculated by [estimated coverage rate change] x [1 + current conversion rate]. In short this is the new conversion rate post-launch that drives the bulk of the ROI calculation.
- Average Order Value: static metric for calculating gross revenue.
- Sessions: static metric for calculating gross revenue.
- Gross Revenue: total revenue from the website (after refunds, cancellations etc), calculated by [total sessions] x [conversion rate ] x [average order value].
- Net Revenue: calculated by [gross revenue] x [gross margin of 40%].
- Net Revenue Difference: calculated by [net revenue] – [baseline current revenue].
- Re-platform Investment: static value of 800 hours x $175/hour.
- Net ROI: calculated by [net revenue difference] – [re-platform investment].
Note: Excluding gross margin in ROI calculations can really change the outcome of an analysis. For every $1 of goods sold, there are hard costs as part of doing business (i.e., manufacturing). So, don’t just take the before/after revenue difference from Google Analytics at face value.
OK, now onto the fun stuff (that is, if you like #s).
Using our assumptions above, here is what our re-platform Net ROI looks like for a 12-month period.
Each row is it’s own unique scenario — from a -20% decrease to a 100% increase in conversion rate.
I know, I know … that’s a lot of #s to look at.
Initially, I’m specifically looking to see when my net ROI trends into a positive # (in this scenario — the $40,000 row) and what estimated conversion rate change this relates to (15% increase).
Here’s the same table highlighted with these values:
Now let’s assume that on day 1 my conversion rate spikes by 15% and sustains this average increase for the entire 12 months (through the ups and downs of seasons).
I can expect to break even after 11 months and won’t begin to see a decent profit until year 2.
The bad: I won’t see a return on my investment until almost an entire year … that’s a lot of time for risks to percolate and change the trajectory of my business.
The good: If conversion rate (and AOV + traffic) remains steady, performing above my previous conversion rate of 0.50%, then my ROI will begin to compound in year 2 and 3.
I know this analysis isn’t perfect and has plenty of holes in it. Part of being a good analyst (like you!) is to minimize tunnel vision and recognize where missing pieces can make your recommendations stronger one way or another.
Initially, I tend to stray away from analysis paralysis (i.e., I don’t need perfection) and use this type of quick analysis to help point me in the right direction without spending a ton of effort.
Here are my own holes I would punch in this analysis:
Hole #1: Why only look at a 12-month window?
A: Unless you are in the $20–$30m+ annual revenue, I just don’t see many businesses accepting a 1–2+ year return on investment for a redesign or re-platform based on the speed at which eCommerce technology is moving today.
In many cases the “platform” (i.e., Shopify, Magento, BigCommerce) isn’t the only significant expense. There are ERPs, CRMs, warehouses, inventory and more changes that require investment but won’t provide immediate tangible revenue growth.
The website needs to (largely) drive revenue and margin.
If you fall in this category — website + ERP + CRM + [xyx] then include these cost and expense changes (e.g., you’re now saving $50,000/year with CRM advancement) into your analysis.
Hole #2: A 0.5% conversion rate is not normal. If you use something more along industry standard like 2 or 3% then the conversion rate change needed is much smaller.
A: This is actually incorrect. If we move the conversion rate up to 2% and decrease sessions so our gross baseline revenue is still $3 million, then a 15% increase in conversion rate produces the same $40,000 net ROI improvement.
Hole #3: Why are you only looking at conversion rate? The new site is going to drive more traffic because of better SEO and content and thus my time to break even will be faster because of the higher # of sessions.
A: This could very well be the case (and I’ve seen this many times), however be cautious in making too many assumptions (e.g. new site is going to boost traffic AND conversion rate) to mitigate against unrealistic expectations for everyone involved in the project.
I suggest including additional scenarios that have a mix of traffic and/or AOV changes so the project plan includes activities (i.e., onsite/offsite SEO, better bundling strategies) that will help you toward these goals.
Hole #4: You aren’t accounting for ongoing maintenance expenses, especially in the replatform scenario, which can drive a significant change in the time to break even.
A: This is a valid point. During a replatform or redesign I would expect ongoing (during the development) costs to continue. Costs to manage your store don’t go away while rebuilding, unless you shut down.
To factor in post launch expense differences, I would add another column into my calculation which would be the expected cost difference for maintenance over a 12 month period. For example — if on Magento 1 you still have hosting charges whereas Shopify or Salesforce Commerce Cloud (aka Demandware) these are associated to a % of revenue.
The point with these “holes” I’m punching are — be open about these scenarios when discussing with your team. It will allow you to hone in on the meat and potatoes of your key decision factors.
I don’t have a concrete blanket stance on which is better — a redesign, re-platform, or upgrade — because each scenario warrants a thorough analysis of your own data and projections.
A 15% increase in conversion rate (from the example above) is really good and would be an achievement to be proud of.
Ultimately what it boils down to is ensuring that your entire team and management are aligned on what your goals are for the project.
- Increasing revenue … through conversion rate or AOV improvements?
- Driving more traffic (which if so, should investment be pivoted more at digital marketing)?
- Reducing costs?
- Enhancing the brand?
- When do you expect a ROI?
Then, once there is alignment on these goals, develop your business case with ROI projections and …
Don’t promise anything or make guarantees.
Assume the worst.
Plan for halfway between the worst and best scenario. (If you assume the worst then hopefully you’re doing something to mitigate this risk!)
Prioritize your budget allocation toward features that will help you achieve your goals.
A/B or user-test when possible.
Assume your plan will change — and when it does, ensure the changes in priorities or effort still align with your goals.
Pull data and report on progress post launch.
And most importantly — don’t be afraid to admit failure when assumptions have gone wrong. It’s better to stop the bleeding early versus waiting for a bad situation to get worse.