klaviyo-performance-tuning

What this skill does

# Klaviyo Performance Tuning

## Overview

Optimize Klaviyo API performance with response caching, request batching, cursor-based pagination, sparse fieldsets, and connection pooling.

## Prerequisites

- `klaviyo-api` SDK installed
- Understanding of Klaviyo's rate limits (75 req/s burst, 700 req/min)
- Redis or in-memory cache (optional)

## Klaviyo API Performance Characteristics

| Operation | Typical Latency | Max Page Size |
|-----------|----------------|---------------|
| Get Profile by ID | 50-150ms | N/A |
| Get Profiles (list) | 100-300ms | 20 (default), 100 (some endpoints) |
| Create Profile | 100-200ms | N/A |
| Create Event | 50-100ms | N/A |
| Get Segment Profiles | 200-500ms | 20 |
| Campaign Operations | 200-500ms | 20 |

## Instructions

### Step 1: Sparse Fieldsets (Reduce Payload Size)

Klaviyo supports JSON:API sparse fieldsets -- request only the fields you need.

```typescript
// BAD: Fetches all profile fields (20+ attributes)
const profiles = await profilesApi.getProfiles();

// GOOD: Only fetch email and firstName (much smaller payload)
const profiles = await profilesApi.getProfiles({
  fieldsProfile: ['email', 'first_name', 'created'],
});
// Note: fieldsProfile uses snake_case field names (API-level names)

// Fetch profiles with included list relationships
const profilesWithLists = await profilesApi.getProfiles({
  fieldsProfile: ['email', 'first_name'],
  include: ['lists'],
  fieldsLists: ['name'],
});
```

### Step 2: Response Caching

```typescript
// src/klaviyo/cache.ts
import { LRUCache } from 'lru-cache';

const cache = new LRUCache<string, any>({
  max: 5000,             // Max cached items
  ttl: 60 * 1000,        // 1 minute default TTL
  updateAgeOnGet: true,   // Reset TTL on access
});

export async function cachedKlaviyoCall<T>(
  key: string,
  fetcher: () => Promise<T>,
  ttlMs?: number
): Promise<T> {
  const cached = cache.get(key);
  if (cached !== undefined) return cached as T;

  const result = await fetcher();
  cache.set(key, result, { ttl: ttlMs });
  return result;
}

// Cache profile lookups by email
export async function getCachedProfile(email: string) {
  return cachedKlaviyoCall(
    `profile:${email}`,
    () => profilesApi.getProfiles({ filter: `equals(email,"${email}")` }),
    5 * 60 * 1000  // 5 minute TTL for profiles
  );
}

// Cache segment data (changes less frequently)
export async function getCachedSegments() {
  return cachedKlaviyoCall(
    'segments:all',
    () => segmentsApi.getSegments(),
    15 * 60 * 1000  // 15 minute TTL for segments
  );
}
```

### Step 3: Efficient Pagination

```typescript
// src/klaviyo/pagination.ts

/**
 * Auto-paginate with configurable page size and concurrency.
 * Klaviyo cursor pagination: response.body.links.next contains the cursor URL.
 */
export async function fetchAllPages<T>(
  fetcher: (pageCursor?: string) => Promise<{
    body: { data: T[]; links?: { next?: string } };
  }>,
  options = { maxPages: 100 }
): Promise<T[]> {
  const results: T[] = [];
  let cursor: string | undefined;
  let pageCount = 0;

  do {
    const response = await fetcher(cursor);
    results.push(...response.body.data);
    pageCount++;

    const nextLink = response.body.links?.next;
    if (nextLink && pageCount < options.maxPages) {
      const url = new URL(nextLink);
      cursor = url.searchParams.get('page[cursor]') || undefined;
    } else {
      cursor = undefined;
    }
  } while (cursor);

  return results;
}

// Usage: fetch all profiles in a list
const allProfiles = await fetchAllPages(
  (cursor) => listsApi.getListProfiles({ id: listId, pageCursor: cursor })
);
console.log(`Total profiles: ${allProfiles.length}`);
```

### Step 4: Request Batching with DataLoader

```typescript
import DataLoader from 'dataloader';
import { ProfilesApi } from 'klaviyo-api';

// Batch profile lookups: multiple getProfile(id) calls within
// a single tick are combined into one getProfiles() call
const profileLoader = new DataLoader<string, any>(
  async (profileIds) => {
    // Klaviyo doesn't have a batch-get-by-IDs endpoint,
    // so we fetch individually but with concurrency control
    const results = await Promise.allSettled(
      profileIds.map(id => profilesApi.getProfile({ id }))
    );
    return results.map(r =>
      r.status === 'fulfilled' ? r.value.body.data : null
    );
  },
  {
    maxBatchSize: 10,  // Stay well under rate limits
    batchScheduleFn: cb => setTimeout(cb, 50),  // 50ms batch window
    cache: true,  // In-memory cache within the DataLoader
  }
);

// These three calls are batched into concurrent requests
const [p1, p2, p3] = await Promise.all([
  profileLoader.load('PROFILE_ID_1'),
  profileLoader.load('PROFILE_ID_2'),
  profileLoader.load('PROFILE_ID_3'),
]);
```

### Step 5: Parallel API Calls with Concurrency Control

```typescript
import PQueue from 'p-queue';

// Process large operations with controlled concurrency
const queue = new PQueue({
  concurrency: 10,    // Max 10 parallel requests
  interval: 1000,     // Per second
  intervalCap: 50,    // 50 requests/second (safe margin under 75 burst)
});

// Example: update 10,000 profiles efficiently
async function bulkUpdateProfiles(updates: Array<{ email: string; data: any }>) {
  let completed = 0;

  const promises = updates.map(update =>
    queue.add(async () => {
      await profilesApi.createOrUpdateProfile({
        data: {
          type: 'profile' as any,
          attributes: { email: update.email, ...update.data },
        },
      });
      completed++;
      if (completed % 500 === 0) {
        console.log(`Progress: ${completed}/${updates.length}`);
      }
    })
  );

  await Promise.allSettled(promises);
  console.log(`Done: ${completed}/${updates.length}`);
}
```

### Step 6: Performance Monitoring

```typescript
// src/klaviyo/perf-monitor.ts

interface PerfMetrics {
  operation: string;
  durationMs: number;
  success: boolean;
  cached: boolean;
}

const perfLog: PerfMetrics[] = [];

export async function measuredCall<T>(
  operation: string,
  fn: () => Promise<T>,
  cached = false
): Promise<T> {
  const start = performance.now();
  try {
    const result = await fn();
    perfLog.push({ operation, durationMs: performance.now() - start, success: true, cached });
    return result;
  } catch (error) {
    perfLog.push({ operation, durationMs: performance.now() - start, success: false, cached });
    throw error;
  }
}

export function getPerfSummary(): Record<string, { avg: number; p95: number; count: number }> {
  const byOp: Record<string, number[]> = {};
  for (const m of perfLog) {
    (byOp[m.operation] ??= []).push(m.durationMs);
  }
  const summary: Record<string, any> = {};
  for (const [op, durations] of Object.entries(byOp)) {
    durations.sort((a, b) => a - b);
    summary[op] = {
      avg: Math.round(durations.reduce((a, b) => a + b, 0) / durations.length),
      p95: Math.round(durations[Math.floor(durations.length * 0.95)]),
      count: durations.length,
    };
  }
  return summary;
}
```

## Error Handling

| Issue | Cause | Solution |
|-------|-------|----------|
| Cache stampede | Many requests on cache miss | Use stale-while-revalidate pattern |
| Pagination timeout | Very large datasets | Set `maxPages` limit, process in chunks |
| Rate limit on bulk ops | Too much concurrency | Reduce `PQueue` concurrency/intervalCap |
| Slow filter queries | Complex filter expressions | Simplify filters, use segment IDs instead |

## Resources

- [Klaviyo API Filtering](https://developers.klaviyo.com/en/reference/api_overview#filtering)
- [JSON:API Sparse Fieldsets](https://jsonapi.org/format/#fetching-sparse-fieldsets)
- [DataLoader](https://github.com/graphql/dataloader)
- [p-queue](https://github.com/sindresorhus/p-queue)

## Next Steps

For cost optimization, see `klaviyo-cost-tuning`.