Part 2: Price Changes & Days on Market — Surface Motivated Sellers Automatically

Use cases at a glance

Price drops and high days-on-market are key signals of seller motivation. A property listed at $500k that drops to $475k after sitting 20 days is broadcasting that the seller wants to move it. An automated price-change agent catches these signals in real time, before other investors notice.

Price change detection agent

To detect price changes, your agent must store historical snapshots of listing data. On each polling cycle, it compares the current snapshot to the previous one, identifying zpids where the price has changed.

Here's the agent configuration:

---
name: price-change-alert
schedule: "0 */4 * * *"  # Every 4 hours
tools:
  - zillow-rapidapi
config:
  search_params:
    zip_codes: ["78704", "78723"]
    property_type: "Single Family"

  price_change_thresholds:
    min_drop_dollars: 5000
    min_drop_pct: 2.0

  zestimate_thresholds:
    gap_pct: 5  # Alert if list price within 5% of Zestimate

  storage:
    type: "json_file"
    path: "/data/price_snapshots.json"
    retention_days: 90

  output:
    channels: ["slack"]
    slack:
      webhook_url: "${SLACK_WEBHOOK_URL}"
---

job: |
  current = await zillow_rapidapi.search(config.search_params)
  previous = load_snapshot(config.storage.path)

  for listing in current:
    zpid = listing.zpid
    if zpid in previous:
      old_price = previous[zpid].price
      new_price = listing.price
      drop_dollars = old_price - new_price
      drop_pct = (drop_dollars / old_price) * 100

      if drop_dollars >= config.price_change_thresholds.min_drop_dollars:
        if drop_pct >= config.price_change_thresholds.min_drop_pct:
          send_price_alert(listing, drop_dollars, drop_pct)

  save_snapshot(current, config.storage.path)

Key mechanics: The agent maintains a price_snapshots.json file mapping zpid to historical price data. On each run, it fetches current listings, compares prices to the snapshot, and alerts only if both thresholds are met (absolute dollar drop AND percentage drop). After processing, it updates the snapshot with current data.

Both thresholds required

Setting both a dollar threshold and a percentage threshold prevents false positives. Requiring both filters out trivial adjustments (a $1 drop on a $500k property = 0.0002%, well below a 2% threshold) while catching genuine price reductions across all price ranges.

Days on Market monitoring

Days on market is a critical signal. A property listed for 45 days in a 30-day-median market suggests overpricing or condition issues. The key is benchmarking against your local median DOM.

A DOM alert doesn't trigger a price change alert. Instead, it flags properties that have been stagnant for too long. Pair this with a price-drop alert: if a property has been on market >60 days and suddenly drops price, that's a high-confidence motivated seller signal.

Zestimate gap analysis

Zillow's Zestimate is an automated valuation model. When a listing price drops to near or below the Zestimate, it often signals the seller has lost confidence in the original asking price and is capitulating toward market reality.

Here's an agent snippet for Zestimate gap tracking:

for listing in current:
  gap_pct = ((listing.zestimate - listing.price) / listing.zestimate) * 100

  if gap_pct >= config.zestimate_thresholds.gap_pct:
    send_zestimate_gap_alert(listing, gap_pct)
    # Interpretation: list price is now within 5% of Zestimate
    # or BELOW Zestimate. Seller is actively discounting.

When a property listed at $550k (Zestimate $550k) has listed at $550k, the gap is 0%. If price drops to $525k, gap becomes 4.5%. Hit 5% and you alert. If it drops to $500k, gap is 9%—strong seller motivation signal.

Weekly stale listing digest

Instead of alerting on every >60-day listing, compile a weekly report showing all stale inventory in your target zips, sorted by age descending.

schedule: "0 9 * * 1"  # Monday 9 AM

job: |
  current = await zillow_rapidapi.search(config.search_params)
  stale = [l for l in current if l.dom >= 60]
  stale = sorted(stale, key=lambda x: x.dom, reverse=True)

  report = format_stale_digest(stale)
  send_slack_message(report)

Sample report output:

HEARTBEAT schedule

Sample alert output

Here's what a price-drop alert looks like in Slack:

FAQ

How do I distinguish a meaningful price drop from a trivial $1,000 reduction?

Set both an absolute threshold (e.g., min $5,000 drop) AND a percentage threshold (e.g., min 2% reduction). Require both to be met. This filters out cosmetic price adjustments and ensures you're seeing genuine seller motivation.

What does a high DOM (days on market) actually indicate?

DOM above the local average suggests overpricing, condition issues, or a motivated seller willing to negotiate. The key is benchmarking against your specific market's median DOM — a 45-day DOM means very different things in San Francisco (hot market) vs. rural Ohio. The agent can pull local median DOM from Zillow data for context.

Can I track price changes on properties I don't own and haven't saved?

Yes — you define your watchlist by zip code, neighborhood, or polygon, not by individual property. The agent monitors all active listings in the target area and alerts on any that match your price-change or DOM thresholds.

How often is the Zillow data updated for price changes?

Price changes typically appear within 2-4 hours of being reported to the MLS, depending on syndication lag. Run your price-change agent every 4 hours to catch most changes within a reasonable timeframe. For very hot markets where timing is critical, run every 2 hours.

Should I alert on price increases?

Typically no—your goal is to find deals, not overpriced inventory. However, if you're tracking a specific property you own or are considering, a price increase could signal rising demand. You could configure a separate alert for price increases on a watchlist of specific zpids, but broad alerts on price increases will add noise.