Routines - Scheduled Tasks in Factors.ai

List all deals or opportunities created in the last 7 days. For each show the account name, domain, deal stage, deal amount where available, owner, and the associated lifecycle stage. Lead with a one-line total count and total value, then a table sorted by deal amount, highest first. Only report. Do not change anything.

Summarize conversion from website visit to form submission for the last 7 days versus the previous 7 days, broken down by marketing channel. Lead with a headline giving the overall conversion rate and its change versus the prior week, then a table of each channel with both weeks and the percentage change. Only report. Do not change anything.

From the '[SEGMENT NAME]' segment, list every account with activity in the last 7 days. For each account show its Engagement Score, Engagement Level, Top Engagement Signals with counts, industry, annual revenue, and how recently it was last active. Rank accounts by Engagement Score, highest first. Sort the accounts into three groups based on their Top Engagement Signals: buying signals present (a pricing page visit, a sign-up or trial start, a demo request, or repeated views of high-intent product or migration content); high engagement with no buying signal yet; and LinkedIn-only, where the sole signal this period is LinkedIn ad engagement. For the top 3 accounts by Engagement Score, also pull the last 30 days of journey: the last 3 pages or pieces of content viewed, and the known contacts with name, email, job title and a short note on what each did. If an account has no known contacts, say so explicitly. Open with a one-paragraph summary covering how many accounts are in scope, how many are Hot, and which account leads the week. Then add a Notable stories section of two to four sentences on each of the three most sales-actionable accounts. Then show the three grouped tables.

For the last 14 days, pull keyword performance by ad group for broad-match and phrase-match keywords, along with the existing negative keyword lists. Recommend per campaign: standard junk negatives to add where the negative list is thin, broad-match keywords that lack phrase or exact variants, and any duplicate or conflicting negatives to resolve. Present as a digest grouped by campaign, listing the recommended negatives in a table the team can review before applying. Only report and recommend. Do not add or remove any keyword.

Create a KPI report on LinkedIn paid clicks, broken down by company name and company domain (only where company name is known) in the last 180 days. Then find me accounts in the '[SEGMENT NAME]' segment. Using both reports, calculate what percentage of accounts in that segment are attributable to each individual campaign. For example: 30% of accounts come from Campaign A, 20% from Campaign C. Attribution rules: if an account was touched by multiple campaigns, count it toward every campaign it engaged with, so percentages may sum to more than 100%. Use the total number of accounts in the segment as the denominator, and separately report how many segment accounts had zero LinkedIn paid clicks so the unattributed portion is visible.

Look at the last 90 days for LinkedIn account [ACCOUNT ID] and tell me: for every change we made, did it help or hurt? Step 1. Pull the change log for currently active campaigns, campaign groups, and ads. For each change note what was changed with name and ID, the change type (budget, bid, targeting, creative, pause/launch), old value to new value, who made it, and when. If nothing changed, say so and stop. Step 2. For each changed object, take an equal-length window right before the change and compare it to the window from the change until now. Show whether impressions, reach, CPM, clicks, CTR, leads, cost per lead, and influenced pipeline moved in the good direction, using arrows. Step 3. Compare each change against account-wide spend and conversions for the same period and against campaign groups we didn't touch, as a control. Step 4. Mark leads and pipeline as too early to tell if the change is recent, and flag anything with too little data or overlapping changes. Say correlated, never caused. Step 5. Order changes by impact, call out the two or three clearest wins and anything that clearly hurt, and end with a short double down on / revisit section. If the change log can't be pulled, stop and say exactly what's missing. Never make up data.

Analyze all active LinkedIn campaigns for LinkedIn Ads account ID [ACCOUNT ID]. Show the impression breakdown by job title and identify where spend is likely being wasted. For each job title include impressions, clicks, CTR, spend, and conversions if available, whether the title seems relevant or potentially irrelevant, and the reason for flagging it. Flag job titles that have high impressions or spend but low or no engagement or conversions, that appear unlikely to be relevant for a B2B buying audience, or that sit outside our typical ICP personas. Rank the flagged job titles by estimated wasted spend and recommend which should be reviewed or excluded. If the data is insufficient to confidently classify a job title, say so clearly instead of making unsupported assumptions.

List every account whose Engagement Level is Hot based on activity in the last 7 days. For each, show company name, domain, Engagement Score, Engagement Level, Top Engagement Signals, and lifecycle stage. Lead with a one-line count of how many accounts are Hot, then the table sorted by Engagement Score, highest first. Only report. Do not change anything.

List companies that visited our website in the last 7 days that do not yet have a lifecycle stage in our CRM, meaning they are not yet tracked as records. For each show company name, domain, Total Sessions, the pages viewed, and Engagement Score. Lead with a one-line count of net-new companies, then a table sorted by Total Sessions, highest first. Only report. Do not change anything.

List the top 15 accounts by Engagement Score based on the last 7 days of activity. For each include company name, domain, Engagement Score, Engagement Level, Top Engagement Signals, and lifecycle stage. Present as a ranked table with a one-line headline naming the single most engaged account. Only report. Do not change anything.

List every account in the [ICP] segment that was active in the last 7 days. For each show company name, domain, Total Sessions, Engagement Score, Engagement Level, and lifecycle stage. Lead with a one-line count of active accounts, then a table sorted by Engagement Score, highest first. Only report. Do not change anything.

Identify the single most engaged account in the [ICP] segment over the last 7 days by Engagement Score, then pull its full account journey. Summarize that account's recent touchpoints, the channels and pages driving engagement, Top Engagement Signals, known contacts with name and job title, and current lifecycle stage. Present as a short briefing a sales rep could read before a call. Only report. Do not change anything.

Create a KPI report on LinkedIn paid clicks, broken down by company name and company domain (only where company name is known) in the last 180 days. Then find me accounts in the '[SEGMENT NAME]' segment. Using both reports, calculate what percentage of accounts in that segment are attributable to each individual campaign. For example: 30% of accounts come from Campaign A, 20% from Campaign C. Attribution rules: if an account was touched by multiple campaigns, count it toward every campaign it engaged with, so percentages may sum to more than 100%. Use the total number of accounts in the segment as the denominator, and separately report how many segment accounts had zero LinkedIn paid clicks so the unattributed portion is visible.

Check daily spend for every active LinkedIn campaign over the last 7 days. Only alert if any single campaign's spend yesterday deviated by more than 40 percent, up or down, from its trailing 7-day average. Skip campaigns whose 7-day average spend is below 25 dollars, since the signal is too noisy. If nothing breaches that, do not notify. When alerting, name each affected campaign with yesterday's spend, its trailing average, and the percentage deviation. Only report. Do not change anything.

Check cost per conversion by Google Ads campaign for the last 7 days versus the prior 7 days, looking only at campaigns with at least 500 dollars of spend in the recent week. Only alert if any qualifying campaign's cost per conversion rose by more than 25 percent. If nothing rose that much, do not notify. When alerting, name the affected campaigns and show a table of their spend, conversions, and cost per conversion for both periods with the percentage change. Only report. Do not change anything.

Check Search Impression Share and Search Budget Lost Impression Share for our brand and high-priority Google Ads campaigns over the last completed week. Only alert if Search Impression Share drops below 90 percent or Search Budget Lost Impression Share exceeds 5 percent. If neither breaches, do not notify. When alerting, for each affected campaign include the impression share, and split the loss into rank-driven versus budget-driven by reporting Search Rank Lost Impression Share and Search Budget Lost Impression Share, with a one-line read of whether to raise budget or investigate competitor pressure. Only report. Do not change anything.