| Course | BU 443 Marketing Analytics |
|---|---|
| Unit | Unit 1 |
| Paper type | Metrics definition exercise |
| Length | About 800 words, 5 pages |
| Format | APA 7 student paper |
| School | Herzing University |
| Program | BS in Marketing |
| Updated | September 2026 |
Free sample paper for BU 443 Unit 1
Two Conversion Rates, Two Return-on-Ad-Spend Figures and One Inflated Open Rate: A Metrics Definition Exercise for a Composite Online Houseplant Retailer
Student Name
Business, Herzing University
BU 443: Marketing Analytics
Instructor Name
January 14, 2026
Two Conversion Rates, Two Return-on-Ad-Spend Figures and One Inflated Open Rate: A Metrics Definition Exercise for a Composite Online Houseplant Retailer
Why Definitions Come First
Greenline Plant Co., a composite online retailer created for this course, sells potted houseplants from a greenhouse in Portland, Oregon and ships them across the continental United States. Its weekly marketing meeting uses about a dozen figures from four systems: the website analytics tool, the store platform that takes orders, the advertising accounts and the email service. In January 2026 the owners noticed that the conversion rate in one report did not match the conversion rate in another, and that the return on social advertising looked twice as good in the ad account as in analytics. Before any analysis, this exercise defines the measures the business reports: what exactly increments each one inside the system that produces it, what its denominator is, and where two figures with the same name are not the same figure.
Farris et al. (2010) argue that many marketing disputes are really disagreements about definitions, and that every metric should come with its formula, its data source and its known distortions. The figures below are from March 2026 for the composite case.
The Measures, Defined
Each measure is defined by the event that increases it, the system that counts it and the denominator used when it is a rate.
| Measure | What increments it | Where counted | Denominator | Known distortion |
|---|---|---|---|---|
| Session | A visit that starts after 30 minutes of inactivity or at midnight | Analytics tool | Not a rate | Visitors who decline analytics cookies are not counted |
| User | A distinct browser identifier seen in the period | Analytics tool | Not a rate | One person on a phone and a laptop counts twice |
| Order | A completed checkout with payment captured | Store platform | Not a rate | Includes orders later refunded |
| Purchaser | A distinct customer email with at least one order | Store platform | Not a rate | Gift buyers with a second email count twice |
| Net sales | Order value less discounts, refunds and replacement credits for damaged plants | Store platform | Not a rate | Refunds land in the month they are issued, not the month of the order |
| Repeat purchase rate | Purchasers whose second order falls within 90 days of their first | Store platform | Purchasers whose first order was at least 90 days ago | Recent customers are excluded until their window closes |
| Email click rate | Unique recipients who clicked any link | Email service | Emails delivered | Link scanners at some companies click automatically |
| Email open rate | An image in the email was loaded | Email service | Emails delivered | Some mail apps load images automatically, so opens are overstated |
Two Conversion Rates
The analytics tool reported a conversion rate of 2.40 percent for March: 3,410 purchase events divided by 142,300 sessions. The store platform reported 2.54 percent: 3,280 sessions that included an order divided by 128,900 sessions. Both are called conversion rate, and neither is wrong.
They differ in both numerator and denominator. The analytics tool counts purchase events, so a customer who places two separate orders in one visit counts twice; the store counts sessions containing at least one order. The analytics tool counts more sessions because it defines a session by inactivity and resets at midnight, while the store uses its own visitor tracking and stops counting visitors who block its script. Kohavi et al. (2020) warn that the same named metric computed in two systems will almost always disagree, and that the choice of which to report must be made once and written down. Greenline will report the store's figure for sales decisions, since it rests on completed orders, and the analytics figure only when comparing traffic sources, since only analytics records where sessions came from.
Two Return-on-Ad-Spend Figures
The social advertising account reported that $9,500 of March spending produced 612 purchases worth $41,800, a return on ad spend of 4.40. The analytics tool credited paid social with 298 purchases worth $19,900, a return of 2.09. The difference comes from how each system decides that an ad caused a purchase. The ad account uses its default attribution setting, crediting a purchase to an ad if the buyer clicked it within seven days or viewed it within one day, and it credits the ad even if the buyer later arrived through email or search. The analytics tool, as configured for this analysis, gives each purchase to the last non-direct source the visitor came from, so a purchase that followed a social ad view but began with a click on an email counts for email.
Neither figure measures what the ads added. Both count purchases that happened near an ad, and some of those buyers would have bought anyway. Li and Kannan (2014) found in a field experiment that rule-based attribution misstated the contribution of channels, and that the incremental effect of a channel can differ considerably from the credit either system assigns. Greenline will report both figures side by side with their rules stated, and treat the gap between them as a reminder that neither is the answer.
An Open Rate That Cannot Be Trusted
The email service reported a 41 percent open rate for March. An open is recorded when a tiny image in the email loads, and some widely used mail apps now load images automatically for privacy reasons, whether or not the person reads the message. The open rate therefore measures, in large part, which mail app subscribers use. Greenline will stop reporting it and report the click rate, 3.1 percent of delivered emails, which requires a person to act, with a note that link-scanning software at some workplaces produces a small number of false clicks.
Two further figures are defined here so they are not confused later. A user is not a customer: one shopper browsing on a phone at lunch and buying on a laptop at night appears as two users and one purchaser. And a purchaser is not a household, since partners who share an address but order under separate emails count separately. Greenline and every figure in this exercise are composites built for the course.
References
Farris, P. W., Bendle, N. T., Pfeifer, P. E., & Reibstein, D. J. (2010). Marketing metrics: The definitive guide to measuring marketing performance (2nd ed.). Pearson Education.
Kohavi, R., Tang, D., & Xu, Y. (2020). Trustworthy online controlled experiments: A practical guide to A/B testing. Cambridge University Press.
Li, H., & Kannan, P. K. (2014). Attributing conversions in a multichannel online marketing environment: An empirical model and a field experiment. Journal of Marketing Research, 51(1), 40-56. https://doi.org/10.1509/jmr.13.0050
BU 443 Unit 1 instructions, in plain terms
The first BU 443 unit asks students to define the marketing measures an organization reports. Prompts typically want several metrics, each with its counting rule inside the system that produces it, its denominator where it is a rate, and its known weaknesses. Many instructors ask students to find two figures that share a name but are calculated differently, such as conversion rate in an analytics tool and in a store platform. Composite data are acceptable. Where a counting rule cannot be found, students should say so and state an assumption. The exercise is usually a definition table followed by short analysis, formatted in APA 7, and it becomes the reference for every later unit in the course.
How this BU 443 Unit 1 example is built
The sample opens with the retailer, its four data systems and the mismatches that prompted the exercise. A definitions table covers eight measures, from sessions to email open rate, each with what increments it, where it is counted, its denominator and its distortion. A section works through two conversion rates, explaining the different numerators and denominators and choosing which to report for which decision. A second section explains the two return-on-ad-spend figures through each system's attribution rules and cites research on incremental effects. A third explains why the open rate is dropped. A short close separates users, customers, purchasers and households.
BU 443 Unit 1 rubric: what full points look like
Rubrics for the metrics exercise at Herzing usually award points for precise definitions, correct identification of counting rules and denominators, recognition of distortions, analysis of conflicting figures and clear presentation. Definitions earn full credit when they state what event increments the measure, not just what it means in general. Denominators matter because most reporting errors come from dividing by the wrong base. Distortion analysis scores highest when specific causes, such as consent or automatic image loading, are named. The conflicting figures criterion rewards explaining both calculations and choosing which to use for which purpose. Instructors also credit students who drop a misleading metric with a reason, rather than reporting it anyway.
BU 443 Unit 1 help before you submit
Metrics exercises often lose points by giving textbook definitions, such as conversion rate is the percentage of visitors who buy, without saying what counts as a visitor or a purchase in a specific system. Another common problem is ignoring denominators, which is where two identical-sounding figures usually diverge. Students frequently treat platform-reported return on ad spend as a measure of what ads caused. Some keep reporting email open rates without acknowledging automatic opens. Others define measures the case never uses while skipping ones it reports every week. Finally, choosing one figure over another without explaining the purpose makes the choice look arbitrary; stating which decision each figure serves resolves that. A useful habit is to keep the definition sheet open during every later unit, since most later arguments turn out to be about which version of a figure was used.
Get a BU 443 Unit 1 example written to your instructions
Send the Unit 1 instructions and the rubric posted in your BU 443 classroom, along with the measures or the case your section assigns. We write a custom example that gives each metric a plain meaning and a counting rule, names its source and denominator, lists exclusions and compares a look-alike pair. First custom sample free, back in 24 to 48 hours.
More BU 443 papers and marketing samples
- BU 443 Unit 2: Data collection and tracking audit
- BU 443 Unit 3: Channel performance analysis
- BU 443 Unit 4: Customer and cohort analysis
- BU 443 Unit 5: Campaign evaluation
- BU 443 Unit 6: Test design and reading
- BU 443 Unit 7: Dashboard or visualization
- BU 443 Unit 8: Analytics recommendation
- BU 231 Unit 1: Global environment overview
- BU 242 Unit 1: Customer value concepts paper
- BU 245 Unit 1: Audience analysis exercise
- BU 302 Unit 1: Organizational setting profile
BU 443 Unit 1 questions, answered
How many measures should the sheet define?
Fewer, defined properly. Six entries carrying counting rules, sources, denominators and exclusions demonstrate more than twenty carrying a sentence apiece, because the criterion is precision rather than coverage. If your instructions fix a count, meet it and spend the remaining effort on the pair you compare, which is the part showing you understood why this exercise exists at all.
What if I cannot find the counting rule for a metric?
Then say so inside the entry, which is a legitimate finding rather than a gap. Write what you could establish, name the documentation you consulted and the date you consulted it, and state which part of the rule remains unknown. A definition marked incomplete stays usable; one that fills the hole with a plausible guess corrupts everything built on it later.
Should I define measures my case does not use?
Only where one is the near twin of a measure it does use, which is exactly the comparison worth making. Otherwise the sheet fills with terms nobody in the case reports and every entry thins out. Define what appears in the reporting your case describes, plus whatever the later units will need, since these definitions are meant to hold across the whole course.
Where can I find a free BU 443 Unit 1 sample paper?
The complete BU 443 Unit 1 metrics definition exercise is on this page for free, with the eight-measure table, both conflicting pairs worked through and margin notes. If your case reports other metrics, we can define them for you in a custom exercise, free and usually inside two days.
Why do two conversion rates differ in BU 443 Unit 1?
They count different things over different bases. In the sample, analytics divides purchase events by its own sessions while the store divides sessions containing an order by its sessions, so both figures are correct for their own rules.