How to save data upload parameters. I'm uploading data from a website daily.

Goal: avoid rebuilding your web query every day by parameterizing the date, turning the query into a reusable function, and (optionally) enabling scheduled refresh/incremental refresh.

Approach A — Parameterize the date & build a reusable function (Power Query)

1. Create a Date parameter
   In Power Query: Home → Manage Parameters → New Parameter → Name DateParam, Type = Date, Current Value = today’s date (or any date to test).
2. Author a query for one date
   Use Get Data → Web to connect once (with any test date). Confirm you can navigate to the table/JSON you need.
3. Convert to a function
   Open the query’s Advanced Editor and parameterize the URL. Prefer Web.Contents with RelativePath/Query so the Service accepts dynamic parts (prevents “dynamic data source” errors).

Example (M) – Stable, refresh-safe pattern

let
    pDate = DateParam,
    BaseUrl = "https://api.example.com",
    DateText = Date.ToText(pDate, "yyyy-MM-dd"),
    Source = Web.Contents(
        BaseUrl,
        [
            RelativePath = "reports/daily",
            Query = [ date = DateText ],
            Headers = [ Accept = "application/json" ]
        ]
    ),
    Json = Json.Document(Source),
    ToTable = Table.FromRecords(Json)
in
    ToTable

Turn it into a function (so you can call it for any date):

(pDate as date) =>
let
    BaseUrl = "https://api.example.com",
    DateText = Date.ToText(pDate, "yyyy-MM-dd"),
    Source = Web.Contents(BaseUrl, [RelativePath = "reports/daily", Query = [date = DateText]]),
    Json   = Json.Document(Source),
    Output = Table.FromRecords(Json)
in
    Output

Now create a simple “runner” query that invokes the function with your parameter, so only the parameter needs updating each day:

let
    TodayData = fxGetDailyData(DateParam)
in
    TodayData

Daily usage: change DateParam in Manage Parameters → Refresh. You never rebuild the query.

Approach B — Auto-pick “today” without manual edits

If the API always needs today’s date, compute it in M:

let
    TodayLocal = Date.From(DateTime.LocalNow()),
    Output = fxGetDailyData(TodayLocal)
in
    Output

Note: Power BI Service evaluates “now” on the service’s region timezone (typically UTC). If the site is date-sensitive to your local time (Europe/London), consider offset logic:

let
    UtcNow     = DateTimeZone.UtcNow(),
    LondonNow  = DateTimeZone.SwitchZone(UtcNow, 0),   // adjust if DST/region differs
    TodayUK    = Date.From(DateTimeZone.RemoveZone(LondonNow)),
    Output     = fxGetDailyData(TodayUK)
in
    Output

Approach C — Append history + Incremental Refresh (no manual daily runs)

If the endpoint can return any past date, store historical rows and refresh only the latest window:

1. Create parameters RangeStart and RangeEnd (Type = DateTime).
2. Make your function accept a Date and filter rows by a date column between those parameters.
3. In the model, right-click the table → Incremental refresh → choose “Store rows in the last N years/months/days” and “Only refresh last N days”.
4. Publish and enable scheduled refresh. Only the recent window is fetched daily.

Example filter step (M)

let
    Raw = fxGetDailyData(Date.From(DateTimeZone.RemoveZone(RangeStart))),
    Filtered = Table.SelectRows(Raw, each [ReportDate] >= Date.From(RangeStart) and [ReportDate] < Date.From(RangeEnd))
in
    Filtered

Approach D — Parameter UI (Date picker) for ad-hoc runs

With a Date parameter, Desktop gives you a date picker. Change it once; the same parameter flows through all dependent queries.

Approach E — Offload the daily download (Power Automate / Dataflow)

• Power Automate: schedule a flow to call the site daily and drop JSON/CSV into SharePoint/OneDrive/Azure Blob. Power BI connects to that folder; each new file appends.
• Power BI Dataflow: build the parameterized query once in a Dataflow, define a parameter at the dataflow level, and schedule the dataflow. Your dataset then reads from the dataflow—no daily Desktop edits.

Important tips & pitfalls

• Dynamic Data Source error: avoid string-concatenated full URLs inside Web.Contents. Use base URL + RelativePath + Query as shown.
• Credentials: set them once (Desktop → publish → Service → Data source credentials). Parameters won’t force you to re-enter unless domain/authority changes.
• Rate limits: if the API throttles, add Retry-After handling (custom function with Function.InvokeAfter), or fetch smaller windows.
• Auditing: stamp the query date into a column so you can distinguish late-arriving corrections vs. original run.

Quick checklist (choose what fits your workflow)

• Desktop-only manual: use Approach A and just change DateParam daily.
• Fully automated in Service: use Approach B (compute today) + scheduled refresh.
• Growing history with fast refresh: use Approach C (Incremental Refresh).
• Enterprise ETL separation: use Approach E (Dataflow/Automate) and point your dataset at curated storage.

Helpful references (verified links)

• Power Query — Using parameters
• Power BI — Configure scheduled refresh
• Power BI — Incremental refresh overview
• Power Query M — Web.Contents
• Power Query documentation hub
• Power BI — Data refresh (concepts & limits)
• Chris Webb — RelativePath & Query options with Web.Contents (expert blog)
• Power Query — Web connector