/
Plan Business Environment

Plan Business Environment

Open Article in Keen Knowledge Base

Background

In the Environment step, you set assumptions about each environment factor during the Plan period. These factors typically have a large impact on overall segment performance so it’s important to be thoughtful in this part of the Plan setup.

See it in action here.

Focus on Volume Impacts

Each environment factor included in your model is outlined with its projected year-over-year sales volume impact during the plan period. These impacts are based on the activity projection type for each factor (default = Keen’s AutoTrend activity).

Pro tip: Work with cross-functional team members to validate and align these values with those coming from your demand planning motions and wargame potential alternatives.

image-20260318-163414.png

Projection Type Baselines

Each of the provided baselines are useful for different applications:

  • AutoTrend - a blend of historical slope and seasonality derived by Keen’s algorithm.   Best broad fit for the widest variety of environment factor types.

  • Match Prior Year – exactly matches the year-over-year activity for each factor, effectively suppressing any environment-driven change.

  • Match Prior Period – exactly matches the period-over-period activity for each factor. This baseline is most appropriate for factors that have some seasonal variability, but the most recent data is representative of your short-term expectations, such as a change in Average Price.

  • Hold to last actualized value – holds the most recent value from your actualized model data flat throughout the plan period. This baseline is most appropriate for factors that have seen historical change that you want to assume will level off going forward, such as Stores Selling.

  • Custom

Baseline Examples

In a Q1 2026 plan built off of a Jan 2023 - Dec 2025 historical model:

  • AutoTrend activity is net new values

    • For trended factors (Ex. Distribution, Price, Category) the AutoTrend produces values using a combination of the rolling average of previous weeks/months, seasonality swells, and historical dataset slope.

    • For on/off factors (Ex. Holidays, Event Markers) activity is marked as either present (1) or not present (0) based on the calendar or historical placement.

      • January: marked with a 1 for all plan weeks that fall in January. All others marked 0.

      • Valentine's Day: This event occurs in one week of the plan, and is marked with a 1.

      • COVID: Historical periods with sales spikes/drops coinciding with COVID stay-at-home will reflect a 1 in the model, and will reflect a 0 (off) in the Plan period.

    • Match Prior Year – activity matches Q1 2025.

    • Match Prior Period – activity matches Q4 2025.

    • Hold to Last Actualized Value – uses the last value for that factor in the historical model. If model runs through week end Dec 31 2025, and the value for Distribution TDPs was 200, every week of the Q1 2026 plan will be 200.

Estimated Year-over-Year Volume Change

Each factor’s estimated volume change is provided as a raw sales volume – just as you see in the decomp chart of your downstream plan simulation results – and as a percentage of total sales volume, to help contextualize the weight of that change.

The calculation itself is based on the percentage of activity change in the period relative to history, multiplied by the factor’s influence (expressed as its mean estimate).

Mean Estimate = the percentage change in sales if your activity changes by 1%.

Volume Change Calculation Examples

In a 13 week Q1 2026 plan built off of a Jan 2023 - Dec 2025 historical model:

  • If Distribution activity for every week in the plan period is expected to be 200 TDPs

  • And Distribution was 180 TDPs in every year-over-year model week (the 13 week span of Q1 2025.)

  • The activity delta between these two periods is (200*13)-(180*13) = 260. This equates to a % change in activity of 260/(180*13) = 11.1%

  • If the Distribution factor has a model mean estimate of 0.50.

  • And the total YOY Sales Volume in the model was 10MM.

  • The impact of this change is estimated to be 0.111*0.50*10MM = 550K

Contextualizing Change

  • A 550K impact relative to an overall starting sales volume of 10MM = (550K/10MM) = 5.5%

  • Across hundreds of user plans, we typically see changes falling within a range of +-15%. Shifts beyond that are atypical. When the activity assumption drives a change that exceeds this threshold, the factor is flagged with a warning icon.

    • Note: changes of larger magnitude are not impossible. Example: a challenger brand who is significantly growing their distribution footprint through new retailers.

Table Edits

Edit the projection baseline for any individual factor from the dropdown, or use the checkboxes to multi-select factors and apply a new baseline globally.

Tip: If you have a lot of environment factors, you can sort the table to surface those that warrant the most careful review.

Factor Modal Live Editing

Click into any individual environment factor to live-edit off of your chosen baseline. The chart reflects both actualized Model data and predicted Plan activity to allow you to better understand the magnitude, shape, and direction of the projection:

  • Hide/Show Historical Values to understand the trendline holistically, or zoom in to clearly see individual week-level projections in the Plan period.

  • Switch to a table view for review or editing of specific values. 

  • Drag individual chart points up or down

  • Multi-Select periods to apply either an absolute value shift or a percentage lift.

The chart and impact result updates instantly with each edit, providing a powerful way to understand the relative weight of each change.

At any point, you can abandon your edits and revert to a precanned baseline – but once you've made edits, the values are considered custom and won't be overwritten unless you explicitly reset them.

image-20260318-171701.png

Upload CSV

If you already have specific activity assumptions in hand, click the “Edit Environment Data” button from the overall table view to download the template with your existing data, edit, and upload it back into the Plan. Existing data will match the AutoTrend values unless a previously edited.

Or, upload the activity for an individual Factor from within its individual modal.

Post-Simulation Comparison

Once a plan is simulated, all volume drivers will be reflected as usual in the decomp chart of the brief.

Beta period note: the default time period in the Brief is prior period. We’re still working to update this default to year-over-year for an easier apples-to-apples comparison with the table in the plan build. In the meantime you can quickly get to the comparison view with a few panel edits:

  • Filter the decomp chart to just the environment factors

  • Switch the tactic view mode from grouped to individual tactics

  • Update the comparison period

Value Comparison

  • The year-over-year impacts you’ll see in the Brief are identical to the upfront values in the plan build for a Status Quo plan.

  • In optimizations, synergystic effects are factored in at plan simulation. As we always say, rising tides lift all boats – so the volume impacts you saw upfront will grow a bit further when the impact of tactic optimization is layered in over top of the starting status qup baseline.

Edge Cases

In models with <52 weeks of historical actuals we’ve used smoothing techniques to fill in year-over-year values where necessary. Use extra diligence when reviewing plans built atop these models.