D381 is MKTG 6040 E-Commerce and Marketing Analytics, worth 3 competency units, and the catalog describes it as selling and reaching customers online together with measuring campaign and website performance. Two subjects in one course, and the join between them is the point: e-commerce is the only marketing environment where the entire path from first impression to completed purchase happens inside systems that record it, so the selling and the measuring are the same subject seen twice.
In e-commerce the funnel is arithmetic, not metaphor
The most useful thing to internalise before writing anything for this course is that an online store's performance decomposes into a handful of multiplied rates, and improving any one of them improves the result predictably.
Sessions times conversion rate gives orders. Orders times average order value gives revenue. Revenue minus cost of goods, fulfilment and acquisition gives contribution. Repeat rate and time between purchases extend that across a customer's life. Every recommendation you make in this course should attach to one of those terms, and saying which one turns a suggestion into an argument.
The decomposition also tells you where to look. A store with plenty of traffic and a low conversion rate has a site or offer problem, not a marketing problem, and spending more on advertising makes it worse by buying more visitors who leave. A store with a healthy conversion rate and thin traffic has the opposite issue. Diagnosing which of those you are looking at, from the numbers, before recommending anything, is often the highest value paragraph in the whole submission.
The second structural point is that cart abandonment is normal rather than scandalous. A large majority of started carts do not complete across the whole industry, so a paper that treats abandonment as a crisis has misread the baseline. What matters is the gap between this store's rate and a comparable benchmark, and what specifically in the checkout produces it.
WGU records the outcome as Competent or Not Competent, with no letter grades and no ordinary grade point average, and performance assessment work can be revised and resubmitted with no penalty attached to the result. Diagnose, recommend, and let the evaluation refine the reasoning.
Turning scored aspects into a store analysis
If your course is assessed by a performance assessment, the aspects the evaluator scores are the outline, each needing a score of 2 on its own.
Here is the arithmetic when data interpretation is involved. Suppose the rubric shows ten scored aspects and the prompt supplies a dataset or a set of performance figures, with a target of about 2,400 words. Data commentary needs room, so allow 400 words for interpreting the supplied numbers, 200 for the business context and 100 for a close, leaving 1,700 across ten aspects, or 170 words each.
Then apply a rule that suits this subject: every recommendation paragraph should contain a number. Not a precise forecast, an order of magnitude. If conversion rate rises from 1.8 percent to 2.2 percent on the same traffic, orders rise by roughly a fifth, and at the stated average order value that is a specific revenue figure. Writing that arithmetic explicitly costs thirty words and converts an opinion into a business case, which is what the aspects in an analytics flavoured course are looking for.
A structure for an e-commerce and analytics deliverable
The usual genre is a performance analysis of an online store with recommendations, sometimes combined with a plan for a new channel or market. This layout keeps diagnosis ahead of prescription.
| Section | What it establishes | Numbers it needs | Where drafts go wrong |
|---|---|---|---|
| Business model | What is sold, to whom, at what margin | Price points and rough unit economics | Recommendations made with no margin in view |
| Traffic analysis | Where visitors come from and how they behave | Sessions by source, bounce and depth | Totals with no breakdown by source |
| Conversion analysis | Where in the path people leave | Step by step drop off through checkout | One overall conversion figure with no funnel |
| Product and merchandising | What sells, what does not, and what is shown first | Revenue concentration across the catalogue | Treating the catalogue as undifferentiated |
| Customer value | What a customer is worth over time | Repeat rate, order frequency, lifetime value estimate | Judging acquisition spend on the first order alone |
| Channel performance | Which acquisition sources pay for themselves | Cost per acquisition against contribution per order | Ranking channels by volume rather than by return |
| Recommendations | What to change, in priority order | Expected effect on a named rate | A long list with no prioritisation |
| Measurement plan | How the changes will be evaluated | Baselines, test design, review cadence | Changes made with no way to attribute the result |
Evidence craft and honest data handling
When a task supplies data, the aspects usually reward how carefully you describe it before interpreting it. Say what the dataset covers, what period, how many records, and what it does not include. A reader who knows the boundaries can trust the analysis inside them.
For external evidence, three families do the work. Platform and analytics documentation defines exactly how each metric is calculated, which matters because a session, a bounce and a conversion are all defined differently across systems. Industry benchmark reports from commerce platforms and payment providers supply conversion, abandonment and average order value ranges, and they come from interested parties, so name the publisher and the year. Government commerce statistics supply the market level context that keeps a store's numbers in perspective.
Two disciplines protect the analysis. Avoid causal language where the design cannot support it: a page changed and revenue rose in the same month is a correlation, and saying so is a strength rather than a hedge. And be careful with averages on skewed data, because a handful of large orders will pull an average order value well above what a typical customer spends. Reporting a median alongside a mean, where the data allows, is a small move that reads as genuine numeric literacy.
What earns Competent, and what comes back
Passing work diagnoses before prescribing, attaches every recommendation to a term in the revenue arithmetic, prioritises by expected effect against effort, and specifies how the change will be measured against a baseline.
Returns cluster in four places. Recommendations with no numbers, so nobody can weigh them. An analysis that reports metrics without ever comparing them to a benchmark or a prior period, which produces description rather than finding. Causal claims from correlational data. And a recommendation list with no order, which leaves an operator no idea what to do first and usually fails an aspect asking for prioritisation.
Prioritisation deserves an explicit method rather than an implied one. The simplest defensible approach is to estimate two things for each recommendation, the expected effect on a named rate and the effort or cost to implement, then rank by the ratio between them. Removing a required account registration from checkout is usually cheap and often moves conversion measurably. Rebuilding a product catalogue is expensive and may move nothing until traffic improves. Writing that comparison as a short table, with your reasoning for each estimate, satisfies the prioritisation aspect directly and takes far less space than arguing the point in prose.
Mobile deserves a specific check in any store analysis. A majority of sessions in most consumer categories arrive on phones while a disproportionate share of completed orders still happen on larger screens, which means the gap between mobile and desktop conversion is often the single largest identifiable loss in the funnel. Segmenting your conversion analysis by device before recommending anything frequently locates the problem in one paragraph, and the fix is usually specific and cheap rather than strategic.
Five mistakes that cost time in D381
Recommending more traffic by default. If conversion is the constraint, more traffic costs money and fixes nothing.
Reporting numbers with no comparison. A figure means nothing without a benchmark, a prior period or a segment to compare against.
Judging channels on volume. The channel that brings the most visitors is often the one that brings the least contribution.
Ignoring repeat purchase. Acquisition economics change entirely once a second order is in view, and lifetime value is usually a scored aspect.
Treating abandonment as failure. Compare against a benchmark and explain the specific friction rather than expressing alarm.
How we work on this course, and where we stop
We work on the written deliverable: mapping the rubric, budgeting aspects to leave room for data commentary, drafting a model analysis with its funnel arithmetic so you can study the method and rewrite it in your own voice, and reading a returned evaluation to name the edits that will clear it. We do not access any store, analytics property or payment system, and we work only from data your prompt supplies or from public sources. If D381 carries an objective assessment on your plan, that exam is proctored and our role is preparation only. We are never present during an assessment, never take one for a student, and never ask for or handle WGU portal credentials.
Analysing store performance for D381?
Send the dataset, the task instructions and the rubric. We come back with the funnel decomposition, the comparisons worth making and a straight read on which recommendations your numbers actually support.
Three questions students ask about D381
How is D381 different from D382 Digital Marketing Analytics?
What do I do if the prompt gives me no dataset?
How do I estimate customer lifetime value at this level?
Where D381 sits in WGU's programs
The July 2026 catalog places this code in 2 current WGU programs. Open a program page for the complete standard path and term positions. The live Degree Plan remains authoritative after transfer credit, substitutions, and mentor planning.
The assessments, one by one
The public catalog does not publish this course's PA/OA identity or task count. WGU Tutors publishes at most one PA manual per course and only from a WGU-controlled public rubric. Until that source exists, PA help begins from the student's real Course of Study and OA support remains preparation only.